CoFe₂O₄ prepared from cobalt-doped coal-gangue-leached iron as an effective peroxymonosulfate activator to degrade rhodamine B: preparation, application and degradation mechanism

Abstract

The typical organic dye rhodamine B (RhB), which poses a serious threat to freshwater safety, could be effectively eliminated via an advanced oxidation technology based on peroxymonosulfate (PMS). Developing efficient catalysts is the key to this process. Herein, CoFe₂O₄ (CFO) catalysts were synthesized via co-precipitation, hydrothermal, and sol–gel methods using an activated acid leaching/extracting/back-extracting process to recover iron from coal gangue as the Fe source and chemically pure cobalt chloride hexahydrate as the cobalt source. The prepared catalysts were applied as a high-efficiency activator of PMS for RhB degradation. As a result, the CFO catalyst prepared by the co-precipitation method could essentially achieve a degradation efficiency of 99.38% for RhB within 20 min; the optimal reaction parameters were a reaction temperature of 20 °C, catalyst concentration of 0.2 g L⁻¹, dye concentration of 30 mg L⁻¹, initial pH of 7, and PMS concentration of 0.7 g L⁻¹. A series of characterizations demonstrated that the catalyst had a large specific surface area, a rich pore structure, and excellent magnetic properties. Combined with radical quenching experiments, EPR and XPS, a possible PMS mechanism was proposed. ¹O₂ was the main active substance for the catalytic system, and SO₄⁻ and ˙OH were also produced during the PMS activation process. The work not only provides a valuable reference for the high-value utilization of high-iron coal gangue but also lays the foundation for the wet leaching of iron from coal gangue to prepare efficient PMS activators for treating organic dye wastewater.

The CFO₁ catalyst activates peroxymonosulfate (PMS) through a bimetallic valence cycle of Co²⁺/Co³⁺ and Fe²⁺/Fe³⁺, efficiently generating reactive oxygen species such as ¹O₂, SO₄˙⁻ and ˙OH to degrade rhodamine B (RhB) pollutants.

1. Introduction

The safety of water is increasingly endangered by the discharge of organic dyes with the rapid development of textile printing and dyeing, biomedicine and other industries in recent years; these issues demand the development of eco-friendly and efficient approaches to treat dye wastewater.^1,2 Advanced oxidation processes (AOPs) are one of the important technologies for purifying dye wastewater through the use of strong oxidative radicals.^3,4 This technique can be used for the oxidative degradation of organic pollutants in various complex environments and includes a range of technical processes, i.e., ozone oxidation, photocatalytic degradation, Fenton degradation, photo-Fenton and electro-Fenton degradation, and UV light degradation.^1,2,5 Peroxymonosulfate (PMS) is one of the most useful oxidants in the field of AOPs because it can be activated to produce SO₄˙⁻ with high redox potential, long half-life, good selectivity, and pH insensitivity, and it can be used for realizing the oxidative degradation of various organic dyes.^6,7 Among many activation methods, various PMS activation technologies based on Fe-based catalysts, such as Fe²⁺, Fe³⁺, Fe₂O₃, MFe₂O₄, etc., have received widespread attention due to their high degradation efficiency, excellent applicability, and low toxicity.^5,8 CoFe₂O₄, as a typical ferrite material with a spinel structure, has been widely used in the field of AOPs due to its excellent magnetic properties, catalytic performance, and extremely high stability.⁸ Conventionally, high-purity analytical-grade iron and cobalt salts (e.g., FeCl₃·6H₂O and Co(NO₃)₂·6H₂O) are used as raw materials for CoFe₂O₄ synthesis via hydrothermal, co-precipitation, or sol–gel routes, which ensures well-controlled crystallinity and uniform active site distribution but inevitably brings about high raw material cost and a large carbon footprint in the reagent production chain.^9–11 In sharp contrast, the synthesis of CoFe₂O₄ from iron-containing solid waste as an alternative iron source avoids the full-cycle resource consumption of industrial iron salt production, and the trace non-ferrous metal impurities naturally retained in the leachate can induce mild lattice distortion during crystallization, which further enriches the oxygen vacancy content on the catalyst surface and imparts unique structural advantages that are difficult to achieve in pure reagent systems.^12,13 However, the synthesis of CoFe₂O₄ using solid waste materials containing Fe and Co, and its AOP performance, needs further research.

Coal gangue is a solid waste from the coal mining process. By 2023, a total of over 6 billion tons of coal gangue had been accumulated. The extensive accumulation of coal gangue has inflicted substantial biological and environmental damage within mining areas.^14,15 Traditional methods to realize the resource utilization of coal gangue, i.e., power generation, backfilling, road construction, and brick making, were widely used from 1960 in China.^16–20 In the past, researchers attempted to use coal gangue to process and produce various zeolites, ceramics, aerogels, and other materials because it contains abundant SiO₂ and Al₂O₃.^21–23 However, coal gangue from certain regions contains larger amounts of Fe₂O₃, suggesting it could be used as an Fe source for the synthesis of spinel ferrites. However, the small amounts of other impurity chemical components significantly impact the synthesis and performance of those materials. Generally, pretreatment of typical high-iron coal gangue by an acid leaching method has been used to achieve the removal of impurity elements and improve the coal gangue grade.^24,25 However, because multiple metal elements in the high-iron coal gangue leached into solution, it was difficult to directly use it as a raw material to produce products via the co-precipitation method. Methods to co-precipitate Fe, Ca, Mg, Al, and other elements to enable recovery and reutilization faced the difficulties of stringent operating conditions, poor selectivity, low separation degree, and impure products when dealing with complex element systems.^26,27 Thus, based on the characteristics of high-iron coal gangue raw materials, finding a reliable method to extract iron from coal gangue, using appropriate methods to prepare catalysts for the treatment of organic pollutants in wastewater, and achieving the goal of “using waste to treat waste” is appealing. This approach is of profound significance for the generation of this type of catalyst and the application of coal gangue.

In this study, the canonical hydrometallurgy process was used to solve the problems mentioned above; high-iron coal gangue was utilized as the raw material to obtain purified Fe ion solutions via the processes of ball milling, high-temperature calcination, HCl leaching, and tributyl phosphate (TBP) back-extraction. Next, Co ions were doped into the obtained purified Fe ion solution. The CoFe₂O₃ catalysts (denoted as CFO catalysts) were synthesized via co-precipitation, hydrothermal, and sol–gel methods. The structure and magnetic properties of the CFO catalysts were characterized using X-ray diffraction (XRD), scanning electron microscopy-energy-dispersive X-ray spectroscopy (SEM-EDS), vibrating sample magnetometer (VSM) studies and Brunauer–Emmett–Teller (BET) analysis. The prepared CFO catalysts were used for PMS activation to degrade RhB in wastewater. The effects of the catalyst preparation method, PMS dosage, and initial pH value on the degradation efficiency of the CFO/PMS system were investigated, and degradation kinetics studies were conducted. The catalytic degradation behavior of RhB molecules was elucidated through radical quenching experiments, X-ray photoelectron spectroscopy (XPS) and electron paramagnetic resonance (EPR) spectroscopy.

2. Experimental section

2.1. Experimental materials

Rhodamine B (C₂₈H₃₁ClN₂O₃, RhB) and methanol (CH₃OH, MeOH) were purchased from Tianjin Kemiou Chemical Co., Ltd (CHN). Cobalt chloride hexahydrate (CoCl₂·6H₂O) was purchased from Meryer (Shanghai) Chemical Technology Co., Ltd (CHN). Citric acid (CA), ammonia (mass ratio of 25%), hydrochloric acid (HCl) and sulfuric acid (H₂SO₄) were purchased from Chongqing Chuandong Chemical (Group) Co., Ltd (CHN). Tributyl phosphate (C₁₂H₂₇O₄P, TBP) was purchased from Aladdin Reagents (Shanghai) Co., Ltd (CHN). Sulfonated kerosene and peroxymonosulfate (KHSO₄·0.5KHSO₄·0.5K₂SO₄, PMS) were purchased from Shanghai Yien Chemical Technology Co., Ltd (CHN). l-Histidine (C₆H₉N₃O₂, l-His) and p-benzoquinone (C₆H₄O₂, BQ) were purchased from Shanghai Macklin Biochemical Co., Ltd (CHN). Sodium thiosulphate (Na₂S₂O₃) and sodium hydroxide (NaOH) were purchased from Tianjin Zhiyuan Chemical Reagent Co., Ltd (CHN). tert-Butyl alcohol (C₄H₁₀O, TBA) was purchased from Tianjin Fuchen Chemical Reagent Co., Ltd (CHN). All chemicals in this work were analytical reagent (AR) and were used directly without further purification.

The coal gangue used in the experiment comes from a Coal Mine in Liupanshui City, Guizhou Province, China, which has the characteristics of a high silicon/aluminum ratio and high iron content. Chemical composition analysis of coal gangue is shown in Table 1.

Table 1. Chemical composition of coal gangue (%).

Composition	SiO2	Al2O3	Fe2O3	TiO2	CaO	K2O	MgO	P2O5	ZrO2	Cr2O3
Content	44.8	17.4	20.6	6.64	3.54	3.11	1.14	0.27	0.218	0.106

2.2. Acid leaching and extraction of Fe from coal gangue

The pretreatment conditions of coal gangue were designed according to previous research.²⁸ The main procedures of this process included calcination, acid leaching, concentration, extraction and back-extraction. The coal gangue was pulverized and ball-milled to pass through a 200-mesh sieve, then calcined in an air atmosphere at 500 °C for 1 h. Afterward, it was leached under conditions of 20% HCl concentration, a leaching temperature of 90 °C, leaching time of 3 h, and liquid–solid mass ratio of 4 : 1 to obtain the coal gangue leaching solution. Next, the leaching solution was collected multiple times and concentrated at 100 °C using a water bath. During this process, the Fe²⁺ in the leaching solution was oxidized to Fe³⁺. After cooling, concentrated HCl was added until the acid concentration of the solution reached more than 15%. TBP was used as the extractant and sulfonated kerosene as the diluent. The extraction was carried out under the conditions of 40% TBP concentration, O/A = 2 : 1 (oil phase/aqueous phase), extraction temperature of 20 °C, and extraction time of 10 min. After phase separation using a separatory funnel, deionized water was used as the back-extraction agent for back-extraction under conditions of O/A = 1 : 1, extraction temperature of 20 °C, and extraction time of 10 min. The above extraction-back extraction process was repeated twice (during which the TBP extractant could be reused); finally, the coal gangue back-extraction solution was obtained.

2.3. Preparation of catalysts

A certain amount of CoCl₂·6H₂O was added to the coal gangue back-extraction solution to make nCo²⁺ : nFe³⁺ = 1 : 2 (molar ratio), and deionized water was added to form a 50 mL homogeneous solution. The CoFe₂O₄ catalysts were then prepared by the co-precipitation method,⁹ hydrothermal method,¹⁰ and sol–gel method.¹¹ Details of the steps are described below.

2.3.1. Co-precipitation method

The solution was heated to 60 °C, and NaOH solution (2 mol L⁻¹) was added dropwise to adjust the pH value to 12 while stirring. Then, the mixed suspension was heated to 80 °C and kept for 1 h. The filtered precipitate was washed with deionized water and ethanol until neutral and dried at 80 °C for 12 h to form a precursor. Finally, the obtained precursor was calcined at 500 °C in an air atmosphere for 2 h to form the catalyst, which was marked as CFO₁.

2.3.2. Hydrothermal method

The solution was heated to 60 °C, and NaOH solution (2 mol L⁻¹) was added dropwise to adjust the pH value to 12 with stirring. The mixed suspension was heated to 80 °C and kept for 1 h, then transferred to a 100 mL Teflon-lined autoclave and kept at 180 °C for 12 h. Next, the black product was filtered and washed with deionized water and ethanol until neutral, dried at 80 °C for 12 h, and finally, the obtained precursor was calcined at 500 °C in an air atmosphere for 2 h to form the catalyst, which was named CFO₂.

2.3.3. Sol–gel method

A certain amount of citric acid (CA) was added to the solution to make nCo²⁺ : nFe³⁺ : nCA = 1 : 2 : 4, then concentrated NH₃·H₂O was added dropwise to the solution to adjust the pH value to 7 at 50 °C with stirring. The mixed solution was then heated to 80 °C and stirred continuously until a gel was formed, and kept at 80 °C for 16 h to form a dry gel. A black fluffy precursor was obtained by heating the dry gel at 200 °C for 2 h. Finally, the precursor was calcined at 500 °C in an air atmosphere for 2 h to form the catalyst, named CFO₃.

2.4. Characterization methods

The phase of the catalyst was characterized with an X-ray powder diffractometer (XRD, Rigaku Ultima IV, Japan), using Cu Kα radiation, with data collection in a continuous scanning angle of 10°–80°. The surface morphology and microstructure of the catalyst were characterized with a scanning electron microscope (SEM, ZEISS Sigma 300, GER) equipped with an energy-dispersive X-ray spectroscopy instrument (EDS, Smartedx, GER). The magnetic properties of the catalyst were tested at ambient temperature with a vibrating sample magnetometer (VSM, LakeShore 7404, USA). The specific surface area, pore size, and pore size distribution of the catalyst were measured with an automatic specific surface and porosity analyzer (Microtrac BEL BELSORP MAX, JPN). X-ray photoelectron spectroscopy (XPS, Thermo Scientific K-Alpha, USA) was used to analyze the elemental composition and valence state of the catalyst. The elemental composition of coal gangue and coal gangue leaching residue was determined by X-ray fluorescence (XRF, Rigaku Supermini 200, JPN) spectroscopy. The concentration of ions in the solution was determined by inductively coupled plasma-optical emission spectroscopy (ICP, Agilent 5110, USA). The UV-visible spectra of the RhB solutions were recorded with a UV-Vis spectrophotometer (TU-1901, CHN). Electron paramagnetic resonance spectroscopy (EPR, Bruker EMXplus-6/1, GER) was used to detect and confirm the active species present during the reaction process.

2.5. Experimental procedures

20 mg of the CFO catalyst was taken into 100 mL of an RhB solution (30 mg L⁻¹) and then placed in a constant-temperature shaker at 20 °C with a speed of 200 r min⁻¹ for 40 min in the dark. After adsorption equilibrium was reached, a certain amount of PMS was added to initiate the reaction. Solutions (0.1 mol L⁻¹) of H₂SO₄ and NaOH were used to adjust the initial pH value of the RhB solution. During the reaction process, a 2 mL sample was taken every 10 min. After separating the catalyst using 0.22 µm filter membranes, 4 mL of 0.063 mol per L⁻¹ Na₂S₂O₃ solution was added to terminate the reaction. The concentration of RhB was measured at a wavelength of 554 nm using a UV-Vis spectrophotometer (TU-1901, CHN). The degradation rate of RhB was calculated using eqn (1) as follows:

1

where R (%) represents the degradation rate, C₀ (mg L⁻¹) represents the initial concentration of RhB, and C (mg L⁻¹) represents the concentration of RhB at a specific time.

After the reaction time reached 60 min, the spent catalyst was recovered using a magnet and used in the next round of cyclic experiments.

3. Results and discussion

3.1. Acid leaching and extraction of coal gangue

The process of utilizing coal gangue for the synthesis of CoFe₂O₄ catalysts is shown in Fig. 1. The entire process includes ball-milling, calcination, acid leaching, extraction, and material preparation. Among them, the ball-milling and calcination steps are particularly important as they are related to whether the mineralogical phase in the coal gangue can be effectively activated, thereby achieving efficient iron removal in the subsequent acid leaching part.^29,30 The chemical composition of the coal gangue leaching slag is shown in Table 2. Through acid leaching, most of the Fe₂O₃, CaO, MgO and other components in the coal gangue were successfully leached, and approximately 85.29% of Fe₂O₃ was leached into the leaching solution. The remaining leaching slag contains more than 80% of the mass ratio of SiO₂ and Al₂O₃, achieving an improvement in the grade of coal gangue, making it suitable as a raw material for producing various silicon-based materials.

Fig. 1. Schematic of the resource utilization of coal gangue.

Table 2. Chemical composition of the leaching residue (%).

Composition	SiO2	Al2O3	Fe2O3	TiO2	CaO	K2O	MgO	P2O5	ZrO2	Cr2O3
Content	70.9	12.0	2.95	9.82	0.143	2.97	0.423	0.0913	0.229	—

It is worth noting that HCl was used in the acid leaching process in this paper. This is because, on the one hand, the use of HCl is universal for the acid leaching treatment of coal gangue and, on the other hand, it creates a suitable anion environment for the subsequent TBP extraction process.³¹ The extraction process of TBP in the HCl system mainly involves the reactions shown in eqn (2) and (3) as follows:³²

Fe³⁺ + 4Cl⁻ ↔ FeCl₄⁻, 2

H⁺ + FeCl₄⁻ + 2TBP ↔ HFeCl₄·2TBP. 3

This is a reversible process, where the acidity is the primary factor influencing the reaction.³³ Therefore, a certain amount of concentrated HCl needs to be added before extraction to ensure the reaction proceeds in the forward direction. Conversely, during back-extraction, the acidity needs to be diminished to facilitate the reaction to proceed in the reverse direction; hence, deionized water is used as the back-extraction agent. The mass ratio of each element in the coal gangue leaching solution and the final back-extraction solution is shown in Table 3. Through the process of ball milling, high-temperature calcination, HCl leaching, and TBP extraction, the Fe in the coal gangue is effectively separated and purified. The Fe content in the back-extraction solution is as high as 99.64%, which can be used as the Fe source for synthesizing CoFe₂O₄.

Table 3. Elemental contents of the leaching and back-extraction solutions (%).

Element	Fe	Al	Ca	Mg	K	Na	P	Mn	Sr	Zn
Leaching solution	46.29	32.16	9.904	6.75	2.84	0.919	0.422	0.243	0.165	0.102
Back-extraction solution	99.64	0.008	0.033	0.003	0.085	0.027	0.084	—	—	0.12

3.2. Phase and textural structure of the CFO catalysts

Fig. 2 shows the XRD diffractogram of CFO catalysts synthesized under different processes. The characteristic peaks at 18.46°, 30.26°, 35.63°,43.14°, 53.79°, 57.10°, and 62.90° can be observed in CFO₁, CFO₂, and CFO₃, corresponding to the (111), (220), (311), (400), (422), (511), and (440) planes, which is consistent with the standard pattern of CoFe₂O₄ (JCPDS: 22-1086), indicating that the coal gangue back-extraction solution can be converted into CoFe₂O₄ material through co-precipitation, hydrothermal, and sol–gel methods.^10,34 Among them, a single CoFe₂O₄ phase peak can be observed in CFO₁ and CFO₂, while additional Co₃O₄ and Fe₂O₃ impurity peaks appear in CFO₃, which is due to the limitations of the sol–gel method in the chloride system. Different from the nitrate system, the chloride system lacks strong oxidizing substances (such as NO₃⁻), which makes it impossible to rapidly decompose and produce the precursor powder with uniform Fe and Co distribution during the self-combustion process of the dry gel, ultimately leading to the formation of impurity phases in the product.³⁵ Meanwhile, the main peak positions of each catalyst shifted significantly towards higher angles, which proves the phenomenon of lattice contraction. The average crystallite sizes of CoFe₂O₄ in CFO₁, CFO₂, and CFO₃ were calculated to be 14.9, 14.7, and 34.6 nm, respectively, using the Scherrer equation and the main diffraction peak (2θ = 35.63°). In summary, high-purity phase CoFe₂O₄ crystals can be obtained from coal gangue leaching solution using co-precipitation and hydrothermal methods, and the crystallite size can be effectively controlled within 20 nm. However, the product obtained by the sol–gel method has impurity phases, and the crystallite size is significantly larger.

Fig. 2. XRD patterns of the CFO catalysts.

The SEM-EDS characterization of the CFO catalysts is shown in Fig. 3. Fig. 3a and b depict the SEM images of CFO₁, while Fig. 3d and e represent CFO₂. Both catalysts exhibit an aggregation morphology formed by the agglomeration of numerous small particles. In CFO₁, the agglomerates are larger, with a rich pore structure on the surface and significant inter-agglomerate pores, contributing to a large specific surface area. In contrast, the degree of agglomeration in CFO₂ is significantly reduced compared with that of CFO₁, with a richer pore structure on the surface and smaller inter-agglomerate pores due to the milder crystallization environment of the hydrothermal process.³⁶ The SEM images of CFO₃ (Fig. 3g and h) differ significantly from those of CFO₁ and CFO₂. CFO₃ is primarily composed of various irregular polyhedral crystals stacked together, with almost no observable pore structure on its surface. According to the XRD characterization results, these crystals are composed of Fe₂O₃, Co₃O₄, and CoFe₂O₄.

Fig. 3. SEM and EDS elemental mapping images of the CFO catalysts: (a–c) CFO₁, (d–f) CFO₂ and (g–i) CFO₃.

The EDS mapping images of CFO₁, CFO₂, and CFO₃ (Fig. 3c, f and i) show that Fe, Co, and O elements are uniformly distributed in CFO₁ and CFO₂, but are extremely unevenly distributed in CFO₃, further indicating the presence of multiple crystal phases in CFO₃. Among the three catalyst materials, the morphological features of CFO₁ and CFO₂ are very similar, with only differences in the degree of agglomeration and pore size distribution. However, CFO₃ appears as a dense granular morphology formed by the stacking of various large crystals, with virtually no pores on its surface.

The specific surface area and pore-size distribution of the catalysts were investigated through N₂ adsorption–desorption experiments, as shown in Fig. 4. The isotherms and pore-size distribution diagrams of CFO₁ and CFO₂ are shown in Fig. 4a and b, respectively. Both CFO₁ and CFO₂ exhibit Type IV adsorption isotherms with obvious H3 hysteresis loops, indicating the presence of mesoporous structures with uneven distribution within the two catalysts.^37,38 The pore-size distribution of CFO₁ and CFO₂ was calculated using the BJH method. As shown in Table 4, CFO₁ has an average pore diameter of 8.59 nm, with pore sizes distributed between 1.0 and 18.4 nm, while CFO₂ has an average pore diameter of 5.59 nm, with pore sizes mainly distributed between 1.0 and 7.0 nm. These data are consistent with the analysis results of the adsorption isotherms. The isotherm of CFO₃ (Fig. 4) indicates that this catalyst exhibits almost no N₂ adsorption performance, suggesting an extreme scarcity of pore channel structures within CFO₃. The BET specific surface areas of CFO₁, CFO₂ and CFO₃ are 65.34 m² g⁻¹, 64.70 m² g⁻¹, and 0.61 m² g⁻¹, respectively. The considerable specific surface areas of CFO₁ and CFO₂ are provided by their rich meso-microporous structures, in contrast to the extremely low specific surface area caused by the dense structure of CFO₃. These results are also consistent with the SEM characterization results.

Fig. 4. N₂ adsorption–desorption isotherms (a) and the corresponding pore size distributions (b) of the CFO catalysts.

Table 4. BET results of CFO₁, CFO₂ and CFO₃.

Catalyst	Surface area (m2 g−1)	Average pore volume (cm3 g−1)	Mean pore size (nm)
CFO1	65.34	8.59	11.37
CFO2	64.70	5.59	6.86
CFO3	0.61	4.62	6.17

The magnetic properties of the CFO catalysts were characterized by VSM, as shown in Fig. 5. All three catalysts exhibit obvious ferromagnetism under the action of an external magnetic field; among them, the saturation magnetization strength of CFO₁ is 64 emu g⁻¹, the remanent magnetization strength is 23.2 emu g⁻¹, and the coercivity is 722.4 O_e. The saturation magnetization strength of CFO₂ is 62.5 emu g⁻¹, the remanent magnetization strength is 22.7 emu g⁻¹, and the coercivity is 701.9 O_e. The magnetic properties of CoFe₂O₄ synthesized via co-precipitation and hydrothermal methods are excellent, with all magnetic properties being similar. In contrast, the saturation magnetization strength of CFO₃ is only 16 emu g⁻¹, which is attributed to the fact that the samples contain Fe₂O₃ and Co₃O₄, which have poor magnetic properties, together with the ferrimagnetic CoFe₂O₄.^38–40 These test results indicated that the catalysts made using the three processes could all be recovered by an external magnet after use.

Fig. 5. Magnetic hysteresis loops of the CFO catalysts.

3.3. Catalytic performance study of the CFO catalysts

3.3.1. Effect of different catalytic systems on the degradation of RhB

The results of the degradation experiments of RhB under different catalyst systems are shown in Fig. 6a. When only catalysts were added, both CFO₁ and CFO₂ reached adsorption equilibrium within 40 min, with an adsorption rate of about 1%, which can be neglected, while CFO₃ basically had no adsorption capacity. When only PMS was added, the degradation rate of RhB was only 17.56% after 1 h. However, when CFO catalysts and PMS were jointly added to the RhB solution, the degradation efficiency of the system significantly increased. Specifically, the CFO₃/PMS system increased the degradation rate of RhB to 49.56%, the CFO₂/PMS system increased the degradation rate of RhB to 87.37%, and the CFO₁/PMS system increased the degradation rate of RhB to 89.31%. It was found that the degradation process of RhB conforms to the first-order kinetic model (kt = −ln(C/C₀)).^38,41 To ensure full consistency of fitting rules and full comparability of kinetic parameters, all first-order kinetic fittings for the four systems in Fig. 6b uniformly adopt the complete 0–60 min time interval covering all sampling points after PMS addition, without excluding any valid experimental data. From the degradation kinetics fitting diagram (Fig. 6b), the CFO₁/PMS system had the fastest degradation rate for RhB; its degradation rate constant was 1.07 times higher than that of the CFO₂/PMS system, 3.46 times higher than that of the CFO₃/PMS system, and 11.33 times higher than that of the PMS system. The introduction of the catalyst significantly augmented the activation efficiency of PMS, leading to a substantial increase in the quantity of strong oxidative free radicals in the solution, thereby greatly enhancing the degradation efficiency of RhB.⁴² The large specific surface area gives CFO₁ and CFO₂ more abundant active sites, resulting in their degradation efficiency being significantly higher than other systems. Moreover, CFO₁ has a broader pore-size distribution and irregular surface, which is conducive to the adsorption and diffusion of RhB molecules, making it more advantageous in the degradation.⁴³ Thus, the subsequent experiments were conducted with CFO₁.

Fig. 6. (a) Degradation curves and (b) degradation kinetics fitting diagram of RhB in different catalytic systems. Experimental conditions: catalyst = 0.2 g L⁻¹, PMS = 0.3 g L⁻¹, RhB = 30 mg L⁻¹, initial pH 5, and temperature = 20 °C.

3.3.2. Effect of different PMS dosages on the degradation of RhB

The effect of different PMS dosages on the degradation efficiency of RhB in the CFO₁/PMS system was investigated (Fig. 7a). The degradation rate of RhB continuously increased with the increase in PMS dosage, reaching 97.87% after 1 h with a PMS dosage of 70 mg. This is attributable to the fact that an increase in PMS dosage leads to a higher concentration of free radicals produced in the solution, resulting in higher degradation efficiency.⁴⁴ However, when the PMS dosage was incrementally increased to 90 mg, the degradation efficiency of the system did not significantly improve, with the degradation rate of RhB only increasing by 0.58% in comparison to a PMS dosage of 70 mg.

Fig. 7. (a) Degradation curves and (b) degradation kinetics fitting diagram of RhB under different PMS dosages. Experimental conditions: catalyst = 0.2 g L⁻¹, RhB = 30 mg L⁻¹, initial pH 5, and temperature = 20 °C.

From the degradation kinetics fitting diagram (Fig. 7b), within the range of 10–70 mg PMS dosage, every increase of 20 mg PMS dosage could gradually increase the reaction rate constant of the system by 149.45%, 32.29%, and 31.50%. When the PMS dosage in the system reached 70 mg, an increase of 20 mg PMS dosage could only increase the reaction rate constant of the system by 7.64%. In the CFO₁/PMS system, when the PMS dosage reached 70 mg, its degradation efficiency was significantly inhibited. This is mainly because the excess HSO₅⁻, SO₄˙⁻, and ˙OH in the solution will produce quenching reactions, producing SO₄²⁻ and HSO₄⁻, with no degradation ability, or SO₅⁻, with relatively weak degradation ability. The specific reaction process is shown in eqn (4)–(6) as follows:⁴²

SO₄˙⁻ + SO₄˙⁻ → 2SO₄²⁻, 4

2˙OH + HSO₅⁻ → SO₄˙⁻ + H₂O + ¹O₂, 5

SO₄˙⁻ + HSO₅⁻ → SO₅⁻ + HSO₄⁻. 6

3.3.3. Effect of different initial pH values on the degradation of RhB

The effect of different initial pH values on the degradation efficiency of RhB in the CFO₁/PMS system is shown in Fig. 8. It was found that the CFO₁/PMS system demonstrated degradation performance within a broad pH range of 3 to 11 (Fig. 8a). The degradation rate of RhB increased with the rise in pH value, reaching an optimum at pH 7, at which pH value the CFO₁/PMS system could essentially achieve complete degradation of RhB within 20 min. The degradation kinetics fitting diagram (Fig. 8b) showed that the degradation rate of the system exhibited an upward trend within the pH range of 3 to 7, with the degradation kinetic constant increasing from 0.02182 min⁻¹ to 0.25438 min⁻¹. However, further increasing the pH value to 11 caused the reaction degradation constant to decrease to 0.00547 min⁻¹. The main reason for the lower degradation efficiency of the CFO₁/PMS system in an acidic environment is that the high concentration of H⁺ will quench the SO₄˙⁻ and ˙OH in the solution (eqn (8) and (9)), where the e⁻ is generated by the decomposition of PMS (eqn (7)) as follows:⁴⁵

HSO₅⁻ → SO₅˙⁻ + H⁺ + e⁻, 7

˙OH + H⁺ + e⁻ → H₂O, 8

SO₄˙⁻ + H⁺ + e⁻ → HSO₄˙⁻. 9

Fig. 8. (a) Degradation curves and (b) degradation kinetics fitting diagram of RhB at different initial pH values. Experimental conditions: catalyst = 0.2 g L⁻¹, PMS = 0.7 g L⁻¹, RhB = 30 mg L⁻¹, and temperature = 20 °C.

As the pH value increases, the concentration of H⁺ in the solution significantly decreases, thereby enhancing the degradation efficiency of the system. When the pH value reached 9, the degradation rate of RhB began to decline. This is because in an alkaline environment, SO₄˙⁻ in the solution will react with H₂O or OH⁻ to produce ˙OH (eqn (10) and (11)) as follows:⁴⁶

SO₄˙⁻ + H₂O → SO₄²⁻ + ˙OH + H⁺, 10

SO₄˙⁻ + OH⁻ → SO₄²⁻ + ˙OH. 11

Although ˙OH is also a strong oxidative free radical, its half-life and redox potential are shorter and lower than those of SO₄˙⁻, leading to a decrease in the degradation efficiency of the catalytic system. Furthermore, when the pH value exceeds 9.4, the form of PMS in the solution will change from HSO₅⁻ to SO₅⁻, which greatly affects the generation of active free radicals.⁴⁷ Therefore, an increase in pH value beyond a certain level is detrimental to the degradation reaction.

Under optimal initial pH and PMS dosage conditions, the synthesized CFO₁ demonstrates a remarkable degradation efficiency of 99.38% for RhB within 20 min. This suggests that CFO₁ possesses significant potential for application in PMS activation for RhB degradation. Compared with similar catalysts reported recently, it was found that the catalytic performance of CFO₁ is not inferior to that of some composite catalytic materials synthesized using chemicals,^48–51 indicating the excellent performance of CFO₁.

3.4. Cycle stability of catalysts

To determine the cyclic stability of the catalyst, CFO₁ was used for five cycles of experiments under the same reaction conditions as in Section 3.3.3. After each reaction, the catalyst was recovered using a magnet, and the concentrations of Fe and Co in the solution were measured. The results, as shown in Fig. 9, indicate that after three cycles of use, the catalytic performance of CFO₁ significantly decreased, with the RhB degradation rate dropping from 100% to 88.63%. This could be due to the active sites on the surface of CFO₁ being occupied by the adsorption of intermediates generated from the degradation of RhB. To restore the performance of the catalyst, based on previous reports on catalyst regeneration processes,⁵² the CFO₁ catalyst used for three cycles was regenerated using the calcination process, i.e., calcination at 400 °C for 2 h in an air atmosphere to remove the intermediates. Next, CFO₁ regained its catalytic activity after calcination, with the degradation rate of RhB increasing from 88.63% back to 100% and maintaining a high degradation rate of 99.65% in the fifth cycle. The results demonstrate that calcination regeneration can completely decompose and remove surface-adsorbed RhB degradation intermediates, without causing collapse of the spinel structure, change in active site valence states, or severe catalyst sintering.⁵³ According to the ICP-MS analysis of the residual solution after the catalytic reaction, the leached concentrations of Fe and Co ions are both stably controlled below 200 µg L⁻¹. This level is far lower than the total cobalt discharge limit of 1000 µg L⁻¹ specified in the Emission Standard of Pollutants for Copper, Nickel and Cobalt Industry (GB 25467-2010) of China,⁵⁴ and the leaching toxicity of the recovered solid CoFe₂O₄ catalyst meets the requirements of the non-hazardous waste identification threshold. The secondary environmental risk of the whole system is therefore at a low and controllable level. CFO₁ exhibited excellent degradation performance and an extremely low ion leaching rate in the cyclic experiments, indicating that it is a catalyst material with good cycle stability.

Fig. 9. Cycle stability test and ion leaching concentration of the CFO₁/PMS system for RhB degradation.

3.5. Mechanism analysis

3.5.1. Verification of the RhB degradation process

The process of RhB degradation was monitored in real time using a UV-Vis absorption spectrometer (Fig. 10), under the conditions of reaction temperature of 20 °C, initial pH of 7, catalyst dosage of 20 mg, PMS dosage of 70 mg, and reaction time of 1 h. RhB, a typical xanthene dye, has two characteristic absorption peaks at 354 and 554 nm in its UV-Vis absorption spectra, which are attributed to the conjugated structure of the RhB molecule.⁵⁵ As the reaction progressed, these two absorption peaks gradually decreased and essentially disappeared after 20 min, indicating that the conjugated structure of the RhB molecule had been completely destroyed.⁵⁶ This phenomenon suggested that the strong oxidative free radicals produced by the CFO₁/PMS system could effectively mineralize organic dyes.

Fig. 10. UV-Vis absorption spectra of an RhB solution.

3.5.2. Identification of active species

In order to ascertain the main active species in the CFO₁/PSM catalyst system, a series of quenching experiments was conducted, utilizing the experimental conditions in Section 3.4 as a control group. Extensive prior research has shown that SO₄˙⁻, ˙OH, ¹O₂, and O₂˙⁻ are crucial free radicals in the CoFe₂O₄/PMS system.^43,57 Therefore, in this work, methanol (MeOH, 500 mmol L⁻¹) was selected as the quencher for SO₄˙⁻ and ˙OH, tert-butanol (TBA, 500 mmol L⁻¹) was selected as the quencher for ˙OH, l-histidine (l-His, 10 mmol L⁻¹) was selected as the quencher for ¹O₂, and benzoquinone (BQ, 5 mmol L⁻¹) was selected as the quencher for O₂^˙−58,59. The results, as depicted in Fig. 11, demonstrated that in the absence of any quencher, the CFO₁/PMS system could degrade 97.48% of RhB dye within 10 min. When MeOH was present in the system, the degradation rate of RhB decreased from 97.48% to 65.73%, and the presence of TBA reduced the degradation rate of RhB from 97.48% to 84.47%. Both MeOH and TBA had a significant inhibitory effect on the degradation efficiency of the catalytic system, confirming the existence of SO₄˙⁻ and ˙OH, and the contribution of SO₄˙⁻ to the catalyst system is significantly greater than that of ˙OH.⁶⁰l-His exerts a substantial inhibitory effect on the degradation efficiency of the catalyst system, which could reduce the degradation rate of RhB from 97.48% to 7.41%, indicating that ¹O₂ is the key active radical in the CFO₁/PSM catalyst system and plays a leading role in the degradation of RhB.⁶¹ The use of BQ resulted in only a slight decrease in the degradation rate of RhB, indicating that the contribution of ˙O₂⁻ to the catalyst system was extremely limited. In summary, the main active species in the CFO₁/PSM catalyst system is ¹O₂, and the roles of SO₄˙⁻ and ˙OH cannot be ignored.

Fig. 11. Free radical quenching experiment.

EPR was utilized to further confirm the active species in the CFO₁/PMS system. As shown in Fig. 12, DMPO (5,5-dimethyl-1-pyrroline N-oxide) was used to trap SO₄˙⁻ and ·OH, and TEMP (2,2,6,6-tetramethylpiperidine-1-oxyl) was used to trap ¹O₂. After PMS was added to the solution containing the catalyst, obvious radical signal peaks can be observed in the EPR spectra. The peaks of DMPO-SO₄^˙− (with six peaks of 1 : 1 : 1 : 1 : 1 : 1) and DMPO–^˙OH (with four peaks of 1 : 2 : 2 : 1) are shown in Fig. 12a, and the peaks of TEMP–¹O₂ (with three peaks of 1 : 1 : 1) are shown in Fig. 12(b).^49,59 This conclusion fully verifies the existence of ¹O₂, SO₄˙⁻ and˙OH in the catalytic process, which is consistent with the conclusion of the quenching experiment.

Fig. 12. EPR spectra of the CFO₁/PMS system using the trapping agent (a) DMPO and (b) TEMP. Conditions: catalyst = 0.2 g L⁻¹, PMS = 0.7 g L⁻¹, reaction time = 10 min, DMPO/TEMP = 100 mM, and reaction temperature = ambient temperature.

3.5.3. XPS analysis

To further investigate the degradation mechanism of the CFO₁/PSM system, XPS was used to analyze the chemical states of Co, Fe, and O elements on the surface of CFO₁ before and after the degradation reaction. Fig. 13 shows the wide-scan XPS spectra of CFO₁. On the surface of CFO₁ both before and after the reaction, obvious peaks of Co, Fe, and O elements can be detected, and the peak intensity has basically not changed, further verifying the excellent chemical stability of this catalyst. In addition, apart from Co, Fe, O and the adventitious contaminant C, a signal peak of Si was also found in the full-spectrum XPS of CFO₁, which may be due to a small amount of Si dissolved from coal gangue.

Fig. 13. Wide-scan XPS spectra of the CFO₁ catalyst before and after the reaction.

The valence state changes of Co, Fe and O in fresh CFO₁ and used CFO₁ were detected through XPS. After participating in the reaction, the chemical states of Co, Fe, and O elements on the CFO₁ surface have undergone significant changes. As shown in Fig. 14(a), the Co 2p_3/2 XPS spectrum of the fresh CFO₁ surface was divided into three peaks at 779.5 eV, 780.9 eV, and 783.3 eV. The former two can be attributed to Co^II in the octahedral position (Oct. Co^II) and Co^II in the tetrahedral position (Tet. Co^II), and the last peak belongs to Co^III in the octahedral position (Oct. Co^III).⁶² In the Co 2p_3/2 XPS spectrum after the CFO₁ reaction (Fig. 14a), the proportion of Oct. Co^II decreased from 37.94% to 35.19%, the proportion of Tet. Co^II decreased from 41.05% to 37.60%, and the proportion of Oct. Co^III increased from 21.01% to 27.21%. As the core electron donor for PMS activation, lattice Co^II at both tetrahedral and octahedral sites donates one electron to HSO₅⁻, directly oxidizing itself to Co^III while generating SO₄^˙−, which explains the overall Co^II-to-Co^III conversion trend observed after the reaction. Notably, the tetrahedral Co^II shows a larger absolute proportion drop (3.45%) than octahedral Co^II (2.75%), confirming that tetrahedral Co^II, with lower coordination constraint and higher electron mobility, acts as the dominant active site for fast single-electron transfer to PMS.

Fig. 14. (a) Co 2p, (b) Fe 2p and (c) O 1s XPS spectra of the CFO₁ catalyst before and after the reaction.

As shown in Fig. 14b, the two peaks at 710.7 eV and 713.4 eV in the Fe 2p_3/2 XPS spectrum of the fresh CFO₁ are attributed to Fe^II and Fe^III,⁶³ respectively. Similar changes also occurred in Fig. 14(b). Before the reaction, the proportion of Fe^II in CFO₁ was 66.88%, and the proportion of Fe^III was 33.12%. When the degradation reaction occurred, the proportion of Fe^II in CFO₁ dropped to 60.67%, and the proportion of Fe^III rose to 39.33%. Since the standard redox potential of Fe^III/Fe^II(0.77 V) is far lower than that of Co^III/Co^II (1.82 V), the Fe^II-to-Fe^III oxidation proceeds at a much slower intrinsic reaction rate, making this step the rate-determining step that governs the overall PMS activation kinetics. The accumulated Fe^III after the reaction further drives the critical intermetallic electron relay: Fe^III spontaneously accepts an electron from the surrounding structural Co^II to regenerate Fe^II, while reducing Co^III back to Co^II, which sustains the continuous redox cycle and avoids rapid passivation of surface active sites during long-term AOP operation. The above changes indicate that the conversion of Co^II/Fe^II to Co^III/Fe^III plays an important role in the degradation reaction of the CFO_1/PMS system.

The O 1s XPS spectrum of the CFO₁ surface is shown in Fig. 14c. Three obvious peaks can be found at binding energies of 529.8 eV, 531.5 eV, and 532.3 eV, which are attributed to lattice oxygen (O_α), surface oxygen (O_β) (due to surface oxygen-containing groups), and adsorbed oxygen (O_γ) (due to surface-adsorbed H₂O), respectively.⁶⁴ Comparing the O 1s XPS spectra before and after the CFO₁ reaction, it was found that the proportion of O_α decreased from 81.07% to 69.74%, which could be ascribed to the redox cycles of Co^III/Co^II and Fe^III/Fe^II. The proportions of O_β in the fresh CFO₁ catalyst increased from 6.28% to 10.91% in the used CFO₁ catalyst, which could be attributed to the reactions of oxygen vacancies (OVs) with PMS and O₂ of the CFO₁ catalyst during the reaction process.^62,63 The proportion of O_γ in the fresh CFO₁ catalyst increased from 12.66% to19.35% in the used CFO₁ catalyst, which could be attributed to the hydroxylation reaction of the CFO₁ catalyst during the reaction process.^63,64 Combined with the O 1s spectra in Fig. 14(c), the proportion of surface lattice oxygen O_α decreased from 81.07% to 69.74%, while the content of the surface-adsorbed oxygen O_β and defect oxygen O_γ increased significantly. This result further confirms that the electron-transfer process on the catalyst surface is not limited to the metal redox pairs: the surface oxygen vacancies generated by lattice oxygen escape also directly participate in the PMS adsorption and electron delivery process, forming a multi-path electron-transfer network that synergistically accelerates the generation of reactive oxygen species. These results indicate that the O component also participated in the reaction. During the reaction process, Fe–O and Co–O bond breakage takes place, and the occurrence of the redox cycles of Co^III/Co^II and Fe^III/Fe^II, and the oxygen vacancies acting as active sites, play an important role. Notably, the significant results are also corroborated by the previous EPR characterization. The rapid generation and presence of various reactive oxygen species (including SO₄˙⁻, ˙OH, and ¹O₂) in the EPR spectra confirm that these pre-existing and reaction-generated oxygen vacancies are stable active sites rather than transient intermediates. The significant increase in O_β and O_γ can be attributed to the surface-adsorbed oxygen-containing intermediates and H₂O, which are produced by the degradation of RhB.^47,65

3.6. Possible catalytic mechanism of PMS activated by CoFe₂O₄ for RhB degradation

Combining the results of radical quenching experiments, EPR and XPS, a possible catalytic mechanism of the CFO₁/PMS system is proposed in Fig. 15. The oxygen vacancies (OVs) can be formed by the escape of oxygen atoms from O_α (eqn (12)). Then the PMS can be activated to produce SO₄˙⁻, SO₅˙⁻ and ˙OH with the assistance of the Co^II/Fe^II, Co^III/Fe^III and OVs (eqn (13)–(15)), thereby realizing the redox cycle of Co^II/Fe^II and Co^III/Fe^III.⁶⁶ During the process, OVs will be transformed into active oxygen (O₀^x) through charge transfer.⁶⁷ The O₀^x formed in the above reaction can generate ¹O₂ with HSO₅⁻ (eqn (16)).⁶⁸ With the production of SO₅^˙−, the self-quenching reaction related to it begins to occur, and the process is accompanied by the production of ¹O₂ and O₂ (eqn (17)–(19)).^48,69 The produced O₂ will be converted into O₂^˙−via the provision of electrons by the OVs on the surface of the CFO₁ (eqn (20)), and O₂^˙− will eventually produce ¹O₂ through eqn (21) and (22).^63,69 In addition, due to the higher oxidation potential of Co^III/Co^II compared to Fe^III/Fe^II, Co^III can be reduced to Co^II by Fe^II, and the heterogeneous chain reaction that is triggered further accelerates the generation of various free radicals (eqn (23)).^41,45

O_α → OVs 12

Co^II/Fe^II + HSO₅⁻ + OVs → Co^III/Fe^III + SO₄^˙− + H⁺ + O₀^x 13

Co^II/Fe^II + HSO₅⁻ + H₂O + OVs → Co^III/Fe^III + SO₄^˙− + 2H⁺ + ^˙OH + O₀^x 14

Co^III/Fe^III + HSO₅⁻ + H⁺ + O₀^x → Co^II/Fe^II + So₅^˙− + H₂O + OVs 15

O₀^x + HSO₅⁻ → ¹O₂ + HSO₄⁻ 16

SO₅^˙− + SO₅^˙− → ¹O₂ + 2SO₄²⁻ 17

SO₅^˙− + SO₅^˙− → O₂ + 2SO₄²⁻ 18

SO₅^˙− + SO₅^˙− → O₂ + S₂O₈²⁻ 19

O₂ + e⁻ → O₂^˙− 20

O₂^˙− + ^˙OH → ¹O₂ + OH⁻ 21

2O₂^˙− + 2H₂O → ¹O₂ + H₂O₂ + 2OH⁻ 22

Co^III + Fe^II → Co^II + Fe^III 23

SO₄²⁻ + ^˙OH + ¹O₂ + RhB → intermediates → CO₂ + H₂O 24

Fig. 15. Catalytic mechanism of the CFO₁/PMS system.

4. Conclusion

This study provides a new feasible scheme for high-value resource recovery of high-iron coal gangue, and realizes the upgrading of coal gangue grade while synthesizing high-performance water-treatment CoFe₂O₄ catalyst materials. The activating—acid leaching—extraction process was used to separate the rich iron from high-iron coal gangue, and the Fe source obtained through back-extraction was used to synthesize the CoFe₂O₄ catalyst for PMS activation to degrade RhB in wastewater. The effects of catalyst preparation methods, PMS dosage, and initial pH value on the degradation efficiency of the CFO/PMS system and PMS activation mechanism were investigated. The Fe content in the obtained back-extraction solution can reach 99.64%. The optimal activation conditions of the CFO/PMS system included a reaction temperature of 20 °C, a catalyst concentration of 0.2 g L⁻¹, a dye concentration of 30 mg L⁻¹, an initial pH of 7, and a PMS concentration of 0.7 g L⁻¹. Cycling experiments show that CFO₁ exhibits excellent regeneration performance, with an extremely low metal ion leaching rate, and could achieve 99.38% degradation of RhB within 20 min. Characterization and degradation experiment results indicate that the catalyst prepared by the co-precipitation method (CFO₁) has a complete and pure CoFe₂O₄ crystal phase, the largest specific surface area, the best magnetic properties, and a rich pore structure. Moreover, its catalytic performance is also the best among the three catalyst materials. Combined with the conclusions of the quenching experiments and EPR and XPS analyses, it is found that ¹O₂ is the main active species in the PMS/CFO₁ catalytic system, and the generation mechanism of each active species in the system is proposed. In short, the work systematically presents methods to obtain Fe from coal gangue and to prepare a CoFe₂O₄ catalyst, and demonstrates that the activated composite of CoFe₂O₄ is an efficient catalyst that can activate PMS to remove RhB in wastewater.

Author contributions

Mingxian Zhang: writing – original draft, data curation, investigation. Ying He: methodology, investigation. Deshun Kong: resources. Jinsong Du: methodology, investigation. Xuemei Zheng: conceptualization, validation, writing – review and editing. Aiyuan Ma: conceptualization, investigation, writing – review and editing, supervision.

Conflicts of interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability

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