A Comparison Study between Wood Flour and Its Derived Biochar for the Enhancement of the Peroxydisulfate Activation Capability of Fe₃O₄

Abstract

In this study, both wood flour (WF) and wood flour-derived biochar (WFB) were used as supports for Fe₃O₄ to activate peroxydisulfate (PDS). The role of different carriers was investigated emphatically from the aspects of catalyst properties, the degradation kinetics of bisphenol A (BPA), the effects of important parameters, and the generation of reactive oxygen species (ROS). Results showed that both WF and WFB could serve as good support for Fe₃O₄, which could control the release of iron into solution and increase the specific surface areas (SSAs). The WFB/Fe₃O₄ had stronger PDS activation capability than WF/Fe₃O₄ mainly due to the larger SSA of WFB/Fe₃O₄ and the PDS activation ability of WFB. Both radical species (•OH and SO₄^•−) and non-radical pathways, including ¹O₂ and high-valent iron-oxo species, contributed to the degradation of BPA in the WFB/Fe₃O₄–PDS process. Moreover, the WFB/Fe₃O₄ catalyst also showed stronger ability to control the iron release, better reusability, and higher BPA mineralization efficiency than WF/Fe₃O₄.

1. Introduction

The pollution of endocrine disrupting chemicals (EDCs) has attracted extensive attention due to the serious danger to the exposed environments. EDCs, also known as environmental hormones, are contaminants that may cause endocrine disorders [1]. The ingestion of EDCs affects the normal function of endogenous hormones, which can further cause adverse health effects on the human body [2]. EDCs accumulate in sewage treatment plant effluents and surface water at levels as high as μg/L, which can pose potential hazard to the safety of drinking water [3].

To decrease the concentration of EDCs in wastewater, sulfate radical (SO₄^•−)-based advanced oxidation technologies (Sr-AOTs) have been applied as effective methods [4,5]. The high oxidation potential (E⁰ = 2.5~3.1 V vs. NHE), wide adaptable pH range (2~8), and relatively long half-life (30–40 μs) make SO₄^•− radicals own a strong oxidizing ability to decompose recalcitrant contaminants [6,7]. SO₄^•− can be produced by the activation of persulfate, including peroxymonosulfate (PMS) and peroxydisulfate (PDS), during which processes the unstable peroxide (-O-O-) bonds are apt to be attacked by electrons or energy for cleavage [8]. PDS was selected as the radical precursor in this study instead of PMS because of its cost effectiveness, high solubility, and chemical stability [9].

Among various PDS activation methods, transition metal possesses the advantages of low energy demand, high efficiency, and mild reaction conditions [10]. Since Fe was regarded as the most efficient, economical, and environmentally friendly metal for PDS activation [11,12], varieties of Fe-based catalysts were developed (e.g., Fe₃O₄/SBA-15 [13], Fe₃C/porous carbon [14]). Fe-based heterogeneous catalysts can well improve some drawbacks of the homogeneous Fe²⁺/PDS process. The application of Fe₃O₄ can not only avoid Fe precipitation at neutral and alkaline conditions, but it can also decrease the self-quenching of precious radicals, and good separation is able to be achieved [15,16]. However, Fe₃O₄ particles easily aggregate during the reaction due to the high surface energy and the interaction induced by magnetic force, which can influence the catalytic capability. Therefore, Fe₃O₄ particles are often immobilized onto various carriers to prevent agglomeration. Several studies have proven that Fe₃O₄ loaded on carriers could improve the catalytic activity and stability. Fe₃O₄ particles could anchor on sepiolite with good dispersion, which was efficient for PDS activation even at alkaline conditions [17]. Fe₃O₄@carbon nanotube composites also exhibited high PDS catalytic activity and a stronger charge transfer ability and oxidation ability [18]. Moreover, carbon material, such as reduced graphene oxide (rGO), has also been found to be a promising support. The rGO-Fe₃O₄ nanoparticle was prepared as a PDS activator for the efficient degradation of trichloroethylene, which could well prevent Fe₃O₄ agglomeration [19]. Although most of these carriers could demonstrate strong stability, they are normally costly and have complex synthetic methods. Therefore, it is necessary to develop cheap and easily prepared carriers to support Fe₃O₄ as an efficient PDS activator.

Biochar has been verified as a promising catalyst carrier due to its low price, high specific surface area, and excellent chemical and biological stability. In addition, biochar alone also has the ability to activate PDS due to the abundant functional groups on the surface [20,21,22]. Moreover, both radical pathways (e.g., hydroxyl radicals (•OH), surface-bound radicals) and nonradical pathways (e.g., singlet oxygen (¹O₂) and direct electron transfer) were identified as the dominant mechanisms [23,24]. The annual yield of waste wood materials from both agriculture and forestry is very high, which is in urgent need for recycling. The wood materials have strong mechanical strength and stability [14], and the wood-derived biochar with rich oxygen-containing functional groups and π-electron density demonstrated excellent performance in PDS activation [25,26]. Although both the wood-based biomass (WBS) and wood-based biochar (WBC) can be potentially good carriers for the Fe₃O₄ catalyst, so far, hardly any studies have been conducted comparing the influence and the role of WBS and WBC on the catalytic performance of Fe₃O₄ for PDS activation.

In this study, both wood flour (WF) and wood flour-derived biochar (WFB) were used as the carriers to support Fe₃O₄ for the activation of PDS. Bisphenol A (BPA) was selected as a typical EDC due to the toxicity and its recalcitrant structure [27,28]. The role of different carriers was investigated emphatically from the aspects of catalyst properties, BPA degradation kinetics, and the generation of reactive oxygen species (ROS). The effects of important parameters were also evaluated. Finally, the BPA degradation mechanisms were proposed.

2. Results and Discussion

2.1. Characterization of the Catalysts

2.1.1. SEM and FESEM

The surface morphologies of WF_x/Fe₃O₄ and WFB_x/Fe₃O₄ were characterized by SEM and FESEM. As shown in Figure S1, the WF had a sheet-like appearance, while after the pyrolysis treatment, the WFB generally retained the sheet-like structure interspersed with some fibers, which is the typical structure of biochar (Figure 1a). Moreover, the size of WFB became smaller than WF, and the surface of WFB became rougher and more irregular. This was probably because high pyrolysis temperature removed lots of the functional groups and destroyed the polymer framework of the pristine WF [29]. Figure 1b shows that the shape of Fe₃O₄ was almost spherical, and the average sizes of Fe₃O₄ nps were found to be ~30 nm. The FESEM images of WF₂₀/Fe₃O₄ and WFB₂₀/Fe₃O₄ shown in Figure 1c–f further demonstrate the successful synthesis of the composites, in which large numbers of Fe₃O₄ nps appeared on the surface of WF₂₀/Fe₃O₄ and WFB₂₀/Fe₃O₄, and the sheet-like structures could not be observed clearly.

2.1.2. XRD

The XRD patterns of Fe₃O₄, WF/Fe₃O₄, and WFB/Fe₃O₄ are shown in Figure 2a,b. The peaks located at 30.2° (220), 35.6° (311), 43.2° (400), 53.8° (422), 57.1° (511), and 62.8° (440) were consistent with the characteristic peaks of Fe₃O₄ (JCPDS No.19-0629) [30]. It was observed that the characteristic peaks were not very sharp, which may be ascribed to the moderate crystallinity of the prepared Fe₃O₄ [31]. After the incorporation of WF or WFB, the obtained WF/Fe₃O₄ and WFB/Fe₃O₄ with different WF or WFB amounts retained the characteristic peaks of Fe₃O₄, which indicated that the prepared WF/Fe₃O₄ and WFB/Fe₃O₄ composites kept the crystal phase of Fe₃O₄. Additionally, the broad diffraction peaks were observed near 24.5° for all WF/Fe₃O₄ and WFB/Fe₃O₄ composites, which could be assigned to the (002) crystal planes of the graphite structure (JCPDS No.75-1621) [32]. It could be obviously seen that the broad diffraction peaks of WFB/Fe₃O₄ were stronger than those of WF/Fe₃O₄, which indicated that the addition of WFB brought a more graphitized structure than that of WF.

The average crystallite size of the Fe₃O₄ nps was calculated with the Debye–Scherrer formula (Equation (1)) [33]. The results are shown in Table S1, and the average crystallite sizes were in the range of 13~16 nm:

$$\mathrm{D}=\mathrm{K}\mathsf{\lambda }/\left(\mathrm{Bcos}\mathsf{\theta }\right)$$

(1)

where D is the crystallite size, K is the constant (0.89), λ is the X-ray wavelength (0.1541841 nm), θ is the Bragg diffraction angle, and B is the full width at half maximum.

Differences in the average grain size of Fe₃O₄ nps were obtained based on FESEM and XRD results, which was probably because the particles observed from FESEM normally consist of several crystal cells and the average crystallite size that could be obtained from XRD should be smaller than that obtained from FESEM [34].

2.1.3. N₂ Adsorption-Desorption

The N₂ adsorption-desorption isotherms and pore size distributions of catalysts are shown in Figure 2c,d. According to the isotherm classification from the International Union of Pure and Applied Chemistry (IUPAC), the isotherms of Fe₃O₄, WF₂₀/Fe₃O₄, and WFB₂₀/Fe₃O₄ belong to type IV with an H1-type hysteresis loop, indicating the existence of typical mesoporous structures of the catalysts [35]. This is consistent with the pore size distribution results. Moreover, compared to Fe₃O₄ (41.32 m²/g), WF₂₀/Fe₃O₄ (76.25 m²/g) and WFB₂₀/Fe₃O₄ (78.65 m²/g) had larger specific surface areas (SSAs) and more micropores, which might provide abundant active sites for catalytic reactions and facilitate the adsorption of the pollutant.

2.1.4. VSM

The magnetic properties of Fe₃O₄, WF₂₀/Fe₃O₄, and WFB₂₀/Fe₃O₄ were assessed by VSM at room temperature (Figure 3). As illustrated, the saturation magnetization (M_s) of Fe₃O₄ was calculated to be 71.58 emu g⁻¹. Although the M_s values of WF₂₀/Fe₃O₄ (67.54 emu g⁻¹) and WFB₂₀/Fe₃O₄ (62.37 emu g⁻¹) were both lower than that of Fe₃O₄ due to the existence of non-magnetic WF or WFB, the two composites still exhibited strong magnetic responses, which ensured their easy recycle via magnetic force without introducing secondary pollution in practical use.

2.1.5. FTIR

FTIR spectra were used to further validate the successful composition of Fe₃O₄ nps with WF or WFB (Figure S2). All the catalysts had characteristic peaks at 3426.87 cm⁻¹ and 1628.81 cm⁻¹, corresponding to the -OH stretching and bending vibration of water, respectively [36]. Meanwhile, the peak intensities of C-H (~2920 cm⁻¹), C=O (1736.27 cm⁻¹), and C-O-C (1052.56 cm⁻¹) vibrations in WFB and WFB₂₀/Fe₃O₄ were weaker than those of WF and WF₂₀/Fe₃O₄ samples, indicating that the oxygen-containing groups in the lignocellulose structures were decomposed and removed during the pyrolysis process [29,37]. In the spectra of Fe₃O₄, WF₂₀/Fe₃O₄, and WFB₂₀/Fe₃O₄, the vibrations of Fe²⁺-O²⁻ (585.33 cm⁻¹) and Fe³⁺-O²⁻ (445.16 cm⁻¹) appeared, implying the formation of Fe₃O₄, which was also consistent with the results of XRD patterns [38]. After the addition of WF or WFB, the band intensities at 2924.59 cm⁻¹ (asymmetric stretching for aliphatic functional groups) and 2856.07 cm⁻¹ (symmetric stretching for aliphatic functional groups) slightly increased [39]. These changes of FTIR peaks proved that the composites of WF/Fe₃O₄ and WFB/Fe₃O₄ were successfully synthesized.

2.2. Evaluation of Catalytic Performance

The degradation of BPA in different processes was evaluated (Figure 4). Before adding PDS, 30 min pre-adsorption was performed to enable the equilibrium of BPA adsorption. As shown in Figure 4a, WF and WFB showed moderate adsorption capacity for BPA, with removal rates of 19.9% and 15.2% within 30 min, while Fe₃O₄ could only adsorb 4.7% of BPA. After compositing with WF or WFB, the adsorption capacities of WF/Fe₃O₄ and WFB/Fe₃O₄ were slightly improved compared to that of Fe₃O₄, which was consistent with the results of SSA. Control experiments were conducted to preclude the direct oxidation of BPA by PDS (Figure 4a). In the WF-PDS process, hardly any BPA could be degraded, which demonstrated that pure WF could not activate PDS to generate effective ROS. In contrast, WFB showed the effective activation of PDS leading to the complete degradation of BPA within 90 min, which might be due to the formation of active sites such as graphitized carbon, oxygen-containing groups, and so on after high-temperature carbonization.

As seen in Figure 4b, 75.5% of BPA was degraded in the Fe₃O₄-PDS system. For catalysts with different amounts of WF, WF₂₀/Fe₃O₄ exhibited excellent catalytic efficiency with a removal rate of 89.7% in 90 min. With more than 20 mg of WF loading, the BPA removal rate decreased, and the BPA removal rate in the WF₂₀₀/Fe₃O₄-PDS process was even lower than that in the Fe₃O₄-PDS process. These results showed that excessive WF loading could decrease the catalytic efficiency of WF/Fe₃O₄. However, the catalytic performance of all the WFB/Fe₃O₄ catalysts was better than that of Fe₃O₄. Although the catalytic efficiency of WFB/Fe₃O₄ slightly decreased as the amount of WFB increased, almost 100% of BPA in these systems was degraded within 50 min.

Table 1. Comparison of the BPA degradation efficiency in different Sr-AOPs.

Sr-AOPs	Conditions	Removal Rate	Main Reactive Species	Ref.
ZnFe2O4-PDS	pH0 = 6, [BPA]0 = 0.1 mM, catalyst 0.5 g/L, [PDS]0 = 5.0 mM	96.5%, 120 min	h+, •OH and SO4•−	[40]
ZnO@AC@FeO-PMS-UV	pH0 = 7, [BPA]0 = 30 mg/L, catalyst 0.4 g/L, [PMS]0 = 4 mM	98.3%, 120 min	1O2, O2•−, •OH and SO4•−	[41]
UVA-LED/CFO-rGO-PMS	pH0 = 7, [BPA]0 = 20 mg/L, catalyst 0.4 g/L, [PMS]0 = 150 mg/L	99.5%, 30 min	•OH and SO4•−	[42]
Fe2+/g-C3N4/LED-PMS	pH0 = 3.5 ± 0.1, [BPA]0 = 0.01 mM, g-C3N4 dosage 0.5 g/L, [Fe2+]0 = 0.01 mM, [PMS]0 = 0.1 mM,	100%, 90 min	O2•−, 1O2, Fe(IV), •OH	[43]
Fe3C/C1000-PMS	pH0 = 3.1, [BPA]0 = 0.1 mM, catalyst 0.1 g/L, [PMS]0 = 2 mM	100%, <30 min	1O2, transferring electrons, •OH and SO4•−	[14]

compared the degradation efficiencies of BPA in different Sr-AOPs. It can be seen that the WFB₂₀/Fe₃O₄-PDS process of the present study could have comparable efficiency toward BPA degradation with other Sr-AOPs. It should also be noted that the same dose of WF and WFB showed different influences on the catalytic performance of Fe₃O₄. To further reveal the roles of WF and WFB in the composites for the activation of PDS, WF₂₀/Fe₃O₄ and WFB₂₀/Fe₃O₄ were selected for the following study.

2.3. Effects of Reaction Parameters on BPA Degradation

2.3.1. Effect of Initial Solution pH

It is well known that solution pH has significant influence on the performance of various AOTs. Herein, the effects of the initial solution pH (pH₀) on BPA degradation in Fe₃O₄-PDS, WF₂₀/Fe₃O₄-PDS, and WFB₂₀/Fe₃O₄-PDS processes were explored (Figure S3 and Figure 5a,b). The BPA removal rate in Fe₃O₄-PDS and WF₂₀/Fe₃O₄-PDS processes decreased from 75.5% to 3.8% and from 88.7% to 5.1%, respectively, as the pH₀ increased from 3.0 to 7.0. By contrast, the WFB₂₀/Fe₃O₄-PDS process showed stronger pH tolerance, and the BPA removal rate decreased from 100% to 11.5% as the pH₀ increased from 3.0 to 7.0. These results indicated that these catalytic processes preferred more acidic conditions for BPA degradation. Under acidic conditions, ferrous ions could more easily enter into the solution from the catalyst surface, and the dissolved Fe²⁺ could activate PDS to accelerate BPA decomposition [44].

2.3.2. Effect of Catalyst Dosage

The effects of WF₂₀/Fe₃O₄ and WFB₂₀/Fe₃O₄ dosages on BPA degradation were investigated (Figure 5c,d). It was found that the increase of WF₂₀/Fe₃O₄ and WFB₂₀/Fe₃O₄ dosages from 0.1 g/L to 1.0 g/L resulted in the significant and continuous promotion of BPA degradation, with the removal rate increasing from 20.9% to 88.7% and from 44.1% to 100%, respectively. This could be explained by the increase of active sites with higher dosages of catalysts to promote the production of ROS. However, when the dosages of WF₂₀/Fe₃O₄ or WFB₂₀/Fe₃O₄ were further increased to 2.5 g/L, the removal rate of BPA decreased slightly. This probably resulted from the self-quenching effects of explosive ROS and the quenching reactions between ROS and PDS when the catalysts were excessive (Equations (2) and (3)) [45]. Moreover, the Fe(II) on the surface of Fe₃O₄ nps also had a strong scavenging effect towards ROS (Equation (4)) [46]. Additionally, excess dosage of catalyst may also increase the mutual magnetic attraction and agglomeration of the catalysts, interfering with their uniform dispersion in solution, which could reduce the surface area of the catalysts and the corresponding catalytic efficiency [47].

$${\mathrm{SO}}_4^{•-}+{\mathrm{SO}}_4^{•-}\to {\mathrm{S}}_2{\mathrm{O}}_8^{2-}{,k=5\times 10}^8{\mathrm{M}}^{-1}{\mathrm{s}}^{-1}$$

(2)

$${\mathrm{SO}}_4^{•-}{+\mathrm{S}}_2{\mathrm{O}}_8^{2-}\to {\mathrm{SO}}_4^{2-}{+\mathrm{S}}_2{\mathrm{O}}_8^{•-},k=5{.5\times 10}^5{\mathrm{M}}^{-1} {\mathrm{s}}^{-1}$$

(3)

$${\mathrm{SO}}_4^{•-}+\mathrm{Fe}\left(\mathrm{II}\right)\to \mathrm{Fe}\left(\mathrm{III}\right){+\mathrm{SO}}_4^{2-}$$

(4)

2.3.3. Effect of PDS Concentration

Generally, the amount of generated ROS directly relates to the PDS concentration. Herein, the effects of PDS concentration on BPA degradation in WF₂₀/Fe₃O₄-PDS and WFB₂₀/Fe₃O₄-PDS processes were examined (Figure 5e,f). It was found that, with the PDS concentration increasing from 1.0 to 5.0 mM, the BPA removal rate increased from 31.2% to 88.7% in the WF₂₀/Fe₃O₄-PDS process and from 61.1% to 100.0% in the WFB₂₀/Fe₃O₄-PDS process, respectively. However, when the PDS concentration further increased to 10.0 mM, the BPA degradation was only improved slightly. The results demonstrated that the excess amount of PDS did not benefit the degradation of target pollutants, mainly because PDS could also act as the ROS quencher competing for the ROS with target pollutants (Equations (2) and (3)).

2.4. Identification of ROS

2.4.1. Quenching Experiments

In order to further understand the different roles of WF and WFB in the evolution of ROS during PDS activation, quenching experiments were conducted in different systems using various quenchers. It is widely accepted that EtOH can rapidly quench both •OH and SO₄^•– (k_•OH = 1.6~7.7 × 10⁷ M⁻¹ s⁻¹, k_SO4•– = 1.2~2.8 × 10⁸ M⁻¹ s⁻¹), while TBA is usually used as an effective quencher for •OH (k_•OH = 3.8~7.6 × 10⁸ M⁻¹ s⁻¹) but is inert towards SO₄^•– (k_SO4•– = 4.0~9.1 × 10⁵ M⁻¹ s⁻¹) [48,49]. As shown in Figure S4 and Figure 6a,b, the inhibiting effect of EtOH on BPA degradation in both systems was significantly stronger than that of TBA. Therefore, both •OH and SO₄^•– should be involved and contribute to BPA degradation, and SO₄^•– may play a more important role than •OH. In addition, ¹O₂ is often the essential ROS in Sr-AOTs, especially for the processes involving biochar as the catalyst. Therefore, FFA was used as the scavenger for ¹O₂ (k_1O2 = 1.2 × 10⁸ M⁻¹ s⁻¹) to examine the possible contribution of ¹O₂ [50]. It can be seen that FFA showed strong inhibition on BPA degradation in both systems, which implied that ¹O₂ possibly contributed to BPA degradation. However, since FFA also reacts with •OH (k_•OH = 1.5 × 10¹⁰ M⁻¹ s⁻¹) [51], the generation of ¹O₂ needs to be further verified by ESR.

2.4.2. ESR Analysis

ESR was performed to confirm the presence of possible ROS in different processes. DMPO was used as the spin trapping agent to identify •OH and SO₄^•–. As shown in Figure 6c, hardly any signals could be detected in the sole-PDS process, indicating that PDS could hardly self-decompose to generate ROS. This phenomenon was consistent with the results of the catalytic degradation experiment (Figure 4a). When WFB and PDS were added together, DMPO-OH (α_N = α_H = 14.9 G) and DMPO-SO₄ (α_N = 13.2 G, α_H = 9.6 G, α_H = 1.48 G and α_H = 0.78 G) signals could be detected, which proved the formation of •OH and SO₄^•–. The above phenomenon might be due to the activation of PDS by the active sites on WFB (Equations (5) and (6)) [52,53]. It was observed that much stronger signals were detected in the Fe₃O₄-PDS, WF₂₀/Fe₃O₄-PDS, and WFB₂₀/Fe₃O₄-PDS systems, which was consistent with the quenching results (Figure S4 and Figure 6a,b). This might be because Fe(II), with high catalytic activity on Fe₃O₄ nps, could have reacted with PDS to generate SO₄^•–, and then the generated SO₄^•– further reacted with H₂O to form •OH (Equations (6) and (7)). Furthermore, TMP was used as the spin trapping agent to trap ¹O₂ (Figure 6d). The TMP-¹O₂ (α_N = 16.9 G) signal could be detected in the WFB₂₀/Fe₃O₄-PDS process, but hardly any signal could be detected in the Fe₃O₄-PDS process, which was probably due to the formation of ¹O₂ between WFB and PDS. Therefore, ¹O₂ should also be the contributor to BPA degradation in the WFB₂₀/Fe₃O₄-PDS process.

$${\mathrm{S}}_2{\mathrm{O}}_8^{2-}{+\mathrm{active}\mathrm{sites}}_{\mathrm{WFB}}\to {\mathrm{SO}}_4^{•-}{+\mathrm{SO}}_4^{2-}{+\mathrm{active}\mathrm{sites}}_{\mathrm{WFB}}^{*}$$

(5)

$${\mathrm{SO}}_4^{•-}{+\mathrm{H}}_2\mathrm{O}\to {\mathrm{H}}^{+}{+\mathrm{SO}}_4^{2-}{+•\mathrm{OH},k=2\times 10}^{-3}{\mathrm{M}}^{-1}{\mathrm{s}}^{-1}$$

(6)

$${\mathrm{S}}_2{\mathrm{O}}_8^{2-}+\mathrm{Fe}\left(\mathrm{II}\right)\to \mathrm{Fe}\left(\mathrm{III}\right)+{\mathrm{SO}}_4^{•-}{+\mathrm{SO}}_4^{2-}$$

(7)

2.4.3. Identification of High-Valent Iron-Oxo Species

Some previous studies have shown that, during the activation of PDS by Fe(II) or Fe(III), corresponding high-valent iron-oxo species (Fe(IV)=O and Fe(V)=O) can form via double-electron transfer, which can further promote the degradation of pollutants [54,55]. Since Fe₃O₄ involves both Fe(II) and Fe(III), both Fe(IV)=O and Fe(V)=O are likely to be generated during the activation of PDS. Moreover, according to some previous results of density functional theory calculation, the oxidation capacity of high-valent iron-oxo species under acidic conditions is stronger than that under alkaline conditions [56], which agrees well with the degradation performance of BPA in WF₂₀/Fe₃O₄-PDS and WFB₂₀/Fe₃O₄-PDS processes, as discussed in Section 2.3.1. Therefore, it is necessary to study the existence of high-valent iron-oxo species in this study.

It is known that both Fe(IV)=O and Fe(V)=O can oxidize PMSO to PMSO₂ through the oxygen atom transfer pathway (Equations (8) and (9)), differing markedly from radical-mediated routes [57]. The degradation of PMSO and the formation of PMSO₂ in different processes are depicted in Figure 7. It can be seen in Figure 7a that PMSO (0.5 mM) could not be oxidized by PDS alone. It was found in Figure 7b–d that 0.057 mM, 0.218 mM, and 0.219 mM of PMSO was degraded within 90 min in Fe₃O₄-PDS, WF₂₀/Fe₃O₄-PDS, and WFB₂₀/Fe₃O₄-PDS processes, respectively. Correspondingly, about 0.047 mM, 0.140 mM, and 0.145 mM of PMSO₂ was generated within 90 min in respective processes, which could well indicate the formation of high-valent iron-oxo species in these three processes. It was also observed that the WFB₂₀/Fe₃O₄-PDS process showed a stronger capability to form high-valent iron-oxo species than the WF₂₀/Fe₃O₄-PDS process. Then, the yield of PMSO₂ (η = Δ[PMSO₂]/Δ[PMSO]) was quantified to evaluate the contribution of high-valent iron-oxo species to PMSO degradation. Figure S5 shows that η values gradually decreased as the reaction proceeded in the Fe₃O₄-PDS process, and it was about 80% at 90 min. Comparatively, the η values kept stable at approximately 65% during the reaction in both WF₂₀/Fe₃O₄-PDS and WFB₂₀/Fe₃O₄-PDS processes. Since the PMSO can also be oxidized by •OH and SO₄^•– via different pathways rather than form PMSO₂ [54], the η value below 100% indicated the presence of •OH and/or SO₄^•–, which was consistent with the results of ESR. Moreover, the majority of PMSO was oxidized to PMSO₂, which also indicated that the high-valent iron-oxo species was the dominant contributor to PMSO degradation.

(8)

(9)

Further investigations were conducted to elucidate the formation mechanisms of high-valent iron-oxo species in Fe₃O₄-PDS, WF₂₀/Fe₃O₄-PDS, and WFB₂₀/Fe₃O₄-PDS processes. As shown in Figure 8a, the dissolved Fe²⁺ and dissolved total irons in Fe₃O₄-PDS, WF₂₀/Fe₃O₄-PDS, and WFB₂₀/Fe₃O₄-PDS processes could be detected, which indicated the presence of homogeneous Fe²⁺ and Fe³⁺ during the catalytic reaction. In addition, with the addition of WF or WFB, the total iron dissolution decreased significantly, and the use of WFB as the support could make more significant control of iron dissolution than that of WF. Previous studies showed that Fe²⁺ could react with PDS to generate Fe(IV)=O under acidic conditions (Equation (10)) [54], but Fe³⁺ could not [58]. Additionally, the surface ≡Fe(II) and ≡Fe(III) may also play important roles in the generation of high-valent iron-oxo species [55]. Some studies assumed that •OH, SO₄^•– and ≡Fe(IV)=O could come from the reaction between surface ≡Fe(II) and PDS (Equations (6), (7) and (11)), while ≡Fe(V)=O could come from the reaction between ≡Fe(III) and PDS (Equation (12)) [59]. To further prove the formation mechanisms of high-valent iron-oxo species, chelating agents (sodium citrate and 1,10-phenanthroline) were added to examine the important roles of surface ≡Fe(II) and ≡Fe(III) of Fe₃O₄ nps. Some studies have reported that ligand exchange could happen between ligands and specific sites of Fe (such as ≡Fe-OH) [55], which could then suppress the capability of catalysts for PDS activation. In other words, 1,10-phenanthroline and sodium citrate could coordinate with Fe(II) and Fe(III) [60], respectively, leading to the inactivation of surface ≡Fe(II) and ≡Fe(III). From Figure 7b–d, it was clearly seen that PMSO degradation and PMSO₂ production were obviously suppressed by the addition of 1,10-phenanthroline and sodium citrate. The results manifested that chelating agents indeed suppressed the capability of surface ≡Fe(II) and ≡Fe(III), leading to a decrease of ≡Fe(IV)=O and ≡Fe(V)=O.

$${\mathrm{Fe}}_{\mathrm{aq}}^{2+}{+\mathrm{S}}_2{\mathrm{O}}_8^{2-}{+\mathrm{H}}_2\mathrm{O}\to {\mathrm{Fe}}_{\mathrm{aq}}^{\mathrm{IV}}{\mathrm{O}}^{2+}{+2\mathrm{SO}}_4^{2-}{+2\mathrm{H}}^{+}$$

(10)

$$\equiv {\mathrm{Fe}\left(\mathrm{II}\right)+\mathrm{S}}_2{\mathrm{O}}_8^{2-}\to \equiv {\mathrm{Fe}\left(\mathrm{IV}\right)=\mathrm{O}+2\mathrm{SO}}_4^{2-}$$

(11)

$$\equiv {\mathrm{Fe}\left(\mathrm{III}\right)+\mathrm{S}}_2{\mathrm{O}}_8^{2-}\to \equiv {\mathrm{Fe}\left(\mathrm{V}\right)=\mathrm{O}+2\mathrm{SO}}_4^{2-}$$

(12)

To sum up, the high-valent iron-oxo species should be mainly produced by two sources: (i) the reaction of dissolved Fe²⁺ with PDS to form Fe(IV)=O; (ii) the reaction of surface ≡Fe(II) and ≡Fe(III) with PDS to form ≡Fe(IV)=O and ≡Fe(V)=O, respectively.

2.4.4. Active Sites Analysis

The valence state of the Fe element in catalysts was investigated by XPS to further examine the contribution of surface ≡Fe(II) and ≡Fe(III) during the catalytic process. Figure 8b–h showed the XPS full survey spectra and the spectra of Fe 2p of different catalysts. As presented in Figure 8b, C (285.6 eV), O (530.2 eV), and Fe (711.8 eV, 724.3 eV) could be detected in these catalysts [61,62]. The Fe 2p spectrum could be deconvolved into six peaks, with 709.86 eV and 723.46 eV corresponding to Fe 2p^3/2, 711.62 eV and 725.22 eV corresponding to Fe 2p_1/2 [63], and the peaks at 718.16 eV and 732.14 eV being the satellite peaks of Fe 2p_3/2 to Fe 2p_1/2 [64], respectively. According to peak intensity, the content of Fe(II) and Fe(III) could be calculated (Table S2). Fe(II) and Fe(III) accounted for 41.56% and 58.44% of total Fe species in Fe₃O₄, 40.00% and 60.00% in WF₂₀/Fe₃O₄, and 37.21% and 62.79% in WFB₂₀/Fe₃O₄, respectively. After the catalytic reaction, the content of Fe(II) decreased to 36.41%, 32.32%, and 32.12%, respectively, and the content ratio of Fe(III) increased in the three catalysts, which indicated that the electrons transferred from Fe(II) to PDS, and thus, the fraction of Fe(III) increased accordingly.

To further study the catalytic mechanism of WFB in WFB₂₀/Fe₃O₄, FT-IR spectra of fresh WFB and the used WFB were compared (Figure S6). After the catalytic reaction, the C=O groups (1736.27 cm⁻¹) almost disappeared, indicating that the C=O groups on WFB were consumed during the catalytic reaction [65]. Thus, it was assumed that the C=O groups were mainly the functional groups contributing to the catalytic activity of WFB, which was consistent with the findings of previous studies that carbon-based materials containing C=O groups, such as quinone or ketone, could activate PDS to produce ¹O₂ (Equations (13)–(15)) [66,67].

$${\mathrm{WFB}=\mathrm{O}+\mathrm{S}}_2{\mathrm{O}}_8^{2-}{+\mathrm{OH}}^{-}\to {\mathrm{WFB}-\mathrm{OH}/\mathrm{SO}}_3{-\mathrm{O}-\mathrm{O}}^{-}{+\mathrm{SO}}_4^{2-}$$

(13)

$${\mathrm{WFB}-\mathrm{OH}/\mathrm{SO}}_3{-\mathrm{O}-\mathrm{O}}^{-}{+\mathrm{OH}}^{-}\to {\mathrm{WFB}-\mathrm{O}}^{-}{/\mathrm{SO}}_3{-\mathrm{O}-\mathrm{O}}^{-}{+\mathrm{H}}_2\mathrm{O}$$

(14)

$${\mathrm{WFB}-\mathrm{O}}^{-}{/\mathrm{SO}}_3{-\mathrm{O}-\mathrm{O}}^{-}{+\mathrm{S}}_2{\mathrm{O}}_8^{2-}{+2\mathrm{OH}}^{-}\to {\mathrm{WFB}=\mathrm{O}+3\mathrm{SO}}_4^{2-}+{}_{}{}^1\mathrm{O}{}_2{+\mathrm{H}}_2\mathrm{O}$$

(15)

2.5. Stability and Reusability of Catalysts

In order to test the stability and reusability of different catalysts, the used catalysts were separated from the reaction solution by a magnet to conduct consecutive catalytic reactions under the same conditions. The separated catalysts were washed with ultrapure water five times, dried in the oven, and directly used for the next cycles. As shown in Figure 9a, the BPA removal rate in the Fe₃O₄-PDS process decreased from 74% in the first cycle to 51% in the 4th cycle. The decrease of the BPA degradation efficiency was probably due to the agglomeration of Fe₃O₄ nps and the reduced content of Fe(II) in it. Compared to Fe₃O₄, both WF₂₀/Fe₃O₄ and WFB₂₀/Fe₃O₄ showed better reusability (Figure 9b,c). Specifically, WFB₂₀/Fe₃O₄ showed the best reusability with almost all BPA being degraded within 90 min after four cycles. These results were closely related to the following synergistic effects between supports (WF or WFB) and the Fe₃O₄ nps: (i) the prevention of agglomeration of Fe₃O₄ nps; (ii) the prevention of leakage of Fe(II). In addition to acting as a carrier, WFB could also react with PDS to form ¹O₂, thus promoting the degradation of BPA.

2.6. Mineralization Performance

The degree of BPA mineralization is also a critical factor to evaluate the catalytic performance of different catalysts. The changes of TOC in WF₂₀/Fe₃O₄-PDS and WFB₂₀/Fe₃O₄-PDS processes are shown in Figure 9d. The TOC removal rate reached 4.6% and 7.2% within 30 min in WF₂₀/Fe₃O₄-PDS and WFB₂₀/Fe₃O₄-PDS processes, respectively, indicating that BPA was more efficiently mineralized in the WFB₂₀/Fe₃O₄-PDS process.

3. Materials and Methods

3.1. Chemicals

BPA (≥99.0%), PDS (K₂S₂O₈, ≥99.0%), FeCl₃·6H₂O (≥99.0%), FeSO₄·7H₂O (≥99.0%), p-benzoquinone (p-BQ, ≥98.0%), furfuryl alcohol (FFA, 98%), and tert-butyl alcohol (TBA, ≥99.0%) were purchased from Sigma-Aldrich, St. Louis, MI, USA. Methanol (MeOH, ≥99.5%), ethanol (EtOH, ≥99.7%), L-histidine, L-ascorbic acid (≥99.7%), CH₃COOH (≥99.5%), CH₃COONa (≥99.0%), sodium citrate (98%), NH₃·H₂O (75%), and 1,10-phenanthroline (≥99.0%) were all purchased from Sinopharm Chemical, Shanghai, China. The methyl phenyl sulfoxide (PMSO, 98.0%), methyl phenyl sulfone (PMSO₂, 98.0%) and 5-dimethyl-1-pyrroline N-oxide (DMPO, 97.0%) were obtained from Macklin, Shanghai, China. The 2, 2, 6, 6-tetramethylpiperidine (TMP, ≥98.0%) was purchased from Shanghai Aladdin Biochemical, Shanghai, China. The dimethyl sulfoxide (DMSO, 99.7%) was obtained from J&K, China. Ultrapure water was used for all experiments and prepared by the HHitech water purification machine.

3.2. Synthesis of the Catalysts

3.2.1. Preparation of WFB

The WFB was prepared by the pyrolysis method. Poplar wood flour with 280–300 mesh size was used as the precursor of WFB, which was obtained from Yixing Wood Flour Factory (Linyi, Shandong Province, China). The wood flour was put into a tubular furnace and heated at 600 °C for 3 h in the N₂ atmosphere with a raping rate of 5 °C/min. After cooling down to room temperature, the black product was obtained as the WFB, which was ground into powders for further use.

3.2.2. Preparation of Fe₃O₄

The coprecipitation method was used to prepare Fe₃O₄ nanoparticles (nps) [68,69]. Briefly, 2.2992 g FeCl₃·6H₂O and 1.1831 g FeSO₄·7H₂O (2:1) were dissolved in 50 mL ultrapure water with N₂ purging in advance, which was heated to 80 °C and maintained for 30 min under mild mechanical agitation. Then, 25 mL of 25% (w/w) NH₃·H₂O was added into the above mixture dropwise until the pH increased to around 10. The mixture was maintained at 80 °C with continuous stirring for 1 h, and N₂ was purged throughout to coprecipitate. The solid phase was separated by the external magnetic field, which was then washed repeatedly with ultrapure water until the liquid supernatant was neutral. Finally, the solid mixture was freeze-dried overnight to obtain the Fe₃O₄ powder.

3.2.3. Preparation of WF/Fe₃O₄ and WFB/Fe₃O₄

WF/Fe₃O₄ and WFB/Fe₃O₄ composites were also prepared by the coprecipitation method. Different amounts (20, 50, 100, and 200 mg) of WF or WFB were dispersed in 50 mL of ultrapure water with the aid of sonication. After full dispersion, N₂ was purged into the mixture to blow out the dissolved oxygen (DO). Then, 2.2992 g FeCl₃·6H₂O and 1.1831 g FeSO₄·7H₂O (2:1) were dissolved in the solution, which was heated to 80 °C and maintained for 30 min under mechanical agitation. The remaining steps were the same as the preparation of Fe₃O₄. The obtained samples with different amounts of WF or WFB were denoted as WF_x/Fe₃O₄ and WFB_x/Fe₃O₄ (x = 20, 50, 100 and 200), respectively.

3.3. Characterization

The morphology of the obtained catalysts was examined by scanning electron microscopy (SEM, FEI Quanta 200, Portland, OR, USA) and field emission scanning electron microscopy (FESEM, Hitachi SU8220, Tokyo, Japan). X-ray diffraction (XRD) patterns of the samples were collected by a Rigaku Ultima IV X-ray diffractometer (Tokyo, Japan) with a Cu-Kα radiation source (1.541841 Å). Scanning rate and 2θ collection range were set at 10°/min and 20~70°, respectively. The functional groups on the surface were identified using a Fourier transforms infrared spectrophotometer (FT-IR, Bruker VERTEX 80 V, Billerica, MA, USA) with the wavelength range of 400~4000 cm⁻¹. Brunauer-Emmett-Teller (BET) specific surface area and pore size distribution were investigated by the nitrogen (N₂) adsorption-desorption method on a Micromeritics ASAP 2020 HD88 instrument (Norcross, GA, USA). The magnetic properties of catalysts were assessed by a vibrating sample magnetometer (VSM, Quantum Design PPMS-9T, San Diego, CA, USA). X-ray photoelectron spectroscopy (XPS, Kratos Axis Ultra DLD, Manchester, UK) was used to analyze the surface chemical composition and element valence states.

3.4. Experimental Conditions

Unless otherwise illustrated, all experiments were carried out in a 50 mL beaker at room temperature under mechanical agitation. In a typical experiment, the catalyst with a final concentration of 1.0 g/L was added and dispersed well in 50 mL of BPA solution, and the reaction was initiated by adding PDS to obtain the final concentration of 5.0 mM. At certain time intervals, 1.0 mL of reaction solution was withdrawn, which was immediately mixed with 0.1 mL of MeOH to terminate the reaction and then filtered through a 0.22 μm membrane for further measurement.

For the experiments of PMSO oxidation, the catalyst was dispersed well into the beaker containing a certain amount of PMSO solution, and the reaction was initiated by adding PDS stock solutions. Samples (1.0 mL) were collected at predetermined time intervals and quickly quenched by DMSO (0.1 mL) and filtered through the 0.22 μm membranes into vials for further measurement.

The initial pH was adjusted with 1.0 M H₂SO₄ or 1.0 M NaOH solution. The batch experiments were conducted in duplicates at least until the errors were below 5%, and the average values obtained were used for plotting.

3.5. Analytical Methods

The BPA concentration was analyzed by High Performance Liquid Chromatography (HPLC; Dionex Ultimate 3000, Sunnyvale, CA, USA) equipped with a reverse-phase C18 column (250 mm × 4.6 mm × 5.0 μm) and an ultraviolet and visible (UV-Vis) spectrophotometry detector, with the detection wavelength set at 225 nm. The mobile phase was a mixture of 70/30% (v/v) methanol and 0.1% phosphoric acid water solution at a flow rate of 1.0 mL/min. The concentrations of PMSO and PMSO₂ were detected by an HPLC equipped with a C18-A column (250 mm × 3.0 mm × 3.0 μm) and a UV-Vis detector at wavelengths of 215 nm. The mobile phase was a mixture of 20/80% (v/v) methanol and 0.1% phosphoric acid water solution at a flow rate of 0.5 mL/min. All of the column temperatures were set at 35 °C. The electron spin resonance (ESR) spectra were obtained by a Bruker EMX-10/12 device with X-band field scanning. The applied instrumental conditions were set as a central magnetic field of 3480 G, resonance frequency of 9.74 GHz, microwave power of 20.00 mW, and sweep time of 30.00 s. The total organic carbon (TOC) concentration was quantified by Analytic Jena multi N/C 3100 TOC. The concentration of total dissolved iron and dissolved Fe²⁺ in the solution was determined by a spectrophotometric method at 510 nm via forming the complex with 1, 10-phenanthroline [43].

4. Conclusions

In this work, WF and WFB were compared as the supports of Fe₃O₄ to enhance the PDS activation performance for the degradation of BPA. Results showed that WFB had more significant control of iron dissolution from Fe₃O₄ than WF. Moreover, WFB/Fe₃O₄ had a stronger PDS activation capability than WF/Fe₃O₄, which was likely due to the larger SSA of WFB/Fe₃O₄ and the PDS activation ability of WFB. Both radicals (•OH and SO₄^•−) and the non-radical pathways, including ¹O₂ and high-valent iron-oxo species, contributed to the degradation of BPA in the WFB/Fe₃O₄-PDS process. In addition, the WFB/Fe₃O₄ also demonstrated better reusability and stronger BPA mineralization performance during the activation of PDS than WF/Fe₃O₄. The use of WFB as the support for Fe₃O₄ may offer a simple and cost-effective option to enhance the PDS activation performance of Fe₃O₄, which can be applied in organic wastewater treatment.