Controllable synthesis of FeMn bimetallic ferrocene-based metal–organic frameworks to boost the catalytic efficiency for removal of organic pollutants

A series of FeMn bimetallic ferrocene-based metal–organic frameworks (FeMn-Fc-MOFs) with various molar ratios of Fe and Mn (1:9, 2:8, 4:6, 6:4) were successfully synthesized using a simple hydrothermal synthesis method and employed as an efficient activator on persulfate (PS) activation for water decontamination. Characterizations demonstrated that Fe and Mn were smoothly introduced into ferrocene-based MOFs and various molar ratios of Fe:Mn had some influence on crystallinity and surface structure of FeMn-Fc-MOFs. Within 120 min, Fe4Mn6-Fc-MOFs demonstrated the best catalytic activity among the different molar ratios, and acid orange 7(AO7) degradation rate was up to 92.0%. In addition, electrochemical experiments revealed that Fe4Mn6-Fc-MOFs possessed superior electron transfer capability than other FeMn-Fc-MOFs, leading to better catalytic performance. Moreover, quenching tests and electron paramagnetic resonance (EPR) detection indicated that hydroxyl radicals and sulfate radicals were both responsible for AO7 decomposition. Notably, the redox cycle of Fe(II)/Fe(III) and Mn(II)/Mn(IV) was discovered in the Fe4Mn6-Fc-MOFs/PS system, which was considered as the limiting process for the cleavage of the O–O bond in PS to generate active radicals. Ultimately, the Fe4Mn6-Fc-MOFs exhibits an excellent universality and good cycling stability for 5 continuous runs. This paper broadens the application of ferrocene-based MOFs on heterogeneous PS activation in environmental catalysis.


Introduction
Advanced oxidation processes (AOPs) have been recognized as promising substitutes for environmental remediation by generating reactive oxygen species (ROSs) such as • OH (1.8-2.7 V) and SO 4 •− (2.5-3.1 V). Compared to • OH-based AOPs, SO 4 •− -based AOPs have a broad application prospect in the decomposition of organic compounds by their higher redox potential, longer half-life, and selectivity (Zeng et al. 2015). As a common kind of oxidant for SO 4 •− generation, persulfate (PS) have drawn much attention in recent years by virtue of its moderate cost and high stability (Laat and Gallard 1999;Liu et al. 2022a,b;Shi et al. 2022;. Generally, UV, electric, thermal, and transition metal ions are used to activate PS to generate SO 4 •− (Furman et al. 2010;J. Li et al. 2018). Among them, the heterogeneous transition metal-based catalytic activation is increasingly favored by researchers due to its efficiency, cleanness, and low energy consumption. Considering the environmental friendliness and cost-effectiveness, ferrous and manganese, as two common transition metals, have been extensively utilized as activators in various catalytic reactions (Rastogi et al. 2009b, a).
Metal organic frameworks (MOFs) are crystalline porous materials formed by self-assembly of inorganic metal centers with organic ligands. The versatility of MOFs, their tunable surface chemical groups (Q. Zhang et al. 2018a, b), high surface area, rich porous structure (H. , and abundant molecular metal positions (Lin et al. 2020;H. Wang et al. 2019;Yu et al. 2019) have attracted great interest in a wide range of research fields. Bimetallic-based catalysts are currently the subject of much research and are employed in a variety of sectors, including electrocatalysis (Zhu et al. 2022), photocatalysis (Pearson et al. 2013), and chemical catalysis (Arif et al. 2018 (Huang et al. 2017), but there are not many applications of bimetallic-based catalysts in environment. Recently, bimetallic MOF materials have been considered as potential candidates for PS or PMS activation due to their high catalytic activity (Y. . The material's potential can be maximized by altering the metal ratio in bimetallic MOFs. Wang et al. produced Fe/ Ti-MOFs (3:1, 1:1, 3:1) using the solvothermal method and applied Fe/Ti-MOFs (3:1) to the photocatalytic degradation of Orange II by activated PS under visible light irradiation (M. Wang et al. 2018). Despite the tunable advantages of bimetallic MOFs, the insufficient intrinsic activity and low stability limit its further application. Ferrocene (Fc) is an effective oxidant with excellent reversible redox properties. However, the direct use of Fc will cause secondary contamination due to recycling difficulties. Zhang et al. modified Fc to MOFs and used it to activate ozone to degrade amaranth dyes (M. W. Zhang et al. 2018a, b). Fc-based MOFs have good thermal stability and are a promising material with a large internal surface area. Lv et al. fabricated three kind of metal-Fc-MOFs (Co, Mn, Fe) using hydrothermal synthesis and found that the Co-Fc-MOFs was most effective at activating PS to pollutant degradation (Lv et al. 2022). However, cobalt was toxic and difficult to use on a large scale. However, there was less studies on dual-metal Fc-MOFs. Therefore, the introduction of Fe and Mn on Fc-MOFs seemed to be a good choice. Dual-metal Fc-MOFs with different transition metals based on ferrocene were fabricated by Sun et al. and demonstrated excellent hydronium storage and magnet performance (R. Sun et al. 2016). The different ratios of Fe and Mn provide additional structural complexity and property tunability. Above all, to the best of our knowledge, FeMn bimetallic ferrocene-based MOFs has not been prepared and investigated in AOPs. Therefore, it is challenging but rewarding to synthesize and explore the catalytic performance of FeMn bimetallic ferrocene-based MOFs with different Fe/Mn ratios.
This work aims to fabricate FeMn-Fc-MOFs using a modified hydrothermal method, explore their physicochemical properties, verify their catalytic performance, and propose a possible mechanism on PS activation. Impressively, the effect of controllable synthesis of MOFs with varying Fe x / Mn y ratios were explored, especially the efficiency of PS activation. In addition, we also evaluated the impacts of PS dosage, catalyst dosage, pH value, and other factors to find the optimal operating parameters, and finally proposed a possible reaction mechanism for PS activation.

Degradation experiments
The catalytic oxidation of AO7 was conducted in aqueous solution to evaluate the catalytic performance of FeMn-Fc-MOFs. A total of five flasks (50 mL) were used for all adsorption and degradation tests at room temperature. For the purpose of creating an adsorption/desorption equilibrium, it is necessary to dissolve FeMn-Fc-MOFs (5 mg) in AO7 solution (20 mL, 20 mg/L) and stir magnetically for 60 min. PS (1.6 mL, 50 mmol/L) was then mixed into the AO7 solution (the initial pH was about 6.7). A tiny portion of the mixture was removed at certain intervals, centrifuged, and the supernatant was retained. The concentration of AO7 was measured using a UV-vis spectrophotometer (China) with an absorbance of 484 nm.
The catalyst was collected as before and washed with distilled water and ethanol three times following AO7 catalytic degradation for the reusability test of FeMn-Fc-MOFs. The resulting catalyst was then dried at 75 °C for recycling. The experimental conditions in this section are the same as above.

Characterization and analysis methods
XRD patterns of the prepared samples were gained with Cu Kα radiation in the range of 2θ = 2°-60° using a D8 diffractometer. Morphological analysis was conducted by scanning electron microscopy (SEM, Zeiss Gemini 300). The chemical state of the materials was measured by XPS and the XPS spectra were documented with a surface science instrument SSX-100. The CHI608C electrochemical analyzer was used for electrochemical analysis of CV and EIS. Electron paramagnetic resonance (EPR) spectra of free radicals were measured on a Bruker-A330.

Characterizations
The crystal structure and phase purity of FeMn-Fc-MOFs with different Fe and Mn molar ratios were determined by XRD analysis. In Fig. 1, XRD patterns of FeMn-Fc-MOFs showed almost the same characteristic peaks (i.e., the 2θ of 6.24, 16.36, 18.88, and 35.52°), indicating similar structural information in the four FeMn-Fc-MOF materials. These peaks showed high crystallinity and corresponded to Fe-Fc-MOFs (Hu et al. 2021). The drop in Mn ratio may be responsible for the decrease in peak intensity at 35.52°. The independent two peaks occur at 9.16° and 30.24° revealing inhomogeneity in morphology of these spherical particles. However, the peak intensity at 35.52° was significantly higher for FeMn-Fc-MOFs with molar ratios of 1:9 and 2:8 than for 4:6 and 6:4, suggesting that the FeMn-Fc-MOFs exhibited somewhat better crystalline at lower Fe/Mn ratios. The broadening peaks assigned at 35.52° of FeMn-Fc-MOFs with molar ratios of 1:9 and 2:8 indicated that the size of the crystallites was quite small and the nanosheets composed of FeMn-Fc-MOFs with molar ratios of 1:9 and 2:8 are very thin. Besides, the small shift in peak location might be explained by partial Fe doping in MOFs ).
The appearance and morphology of the obtained samples were examined using SEM. The SEM picture in Fig. 2 clearly revealed that the produced samples were hollow microspheres with holes in the shell formed of nanosheets, and an average diameter about 8 μm. It can be seen in Fig. 2a-d that the surface of the microspheres evolved from loose to texture-dense as the proportion of Fe increased, which may be due to the increase in the proportion of Fe causing more Fe and Mn loading on the microspheres. It is generally convinced that such a porous structure would provide a greater amount of active sites for chemical reactions (Yi et al. 2019). Figure 3a displays the removal efficiency of AO7 in FeMn-Fc-MOFs/PS system with various Fe and Mn molar ratios. As observed, after attaining adsorption equilibrium, the adsorption efficiencies of Mn-Fc-MOFs, Fe 1 Mn 9 -Fc-MOF, Fe 2 Mn 8 -Fc-MOFs, and Fe 6 Mn 4 -Fc-MOFs were 5%, 20%, 25%, and 33%, respectively. Only 53% of AO7 remained in Fe 4 Mn 6 -Fc-MOFs system, owing primarily to adsorption. As a result, Fe 4 Mn 6 -Fc-MOFs demonstrated superior adsorption capacity. Fe 1 Mn 9 -Fc-MOFs and Fe 2 Mn 8 -Fc-MOFs achieved 85% and 77% removal of AO7 within 60 min by activation of persulfate, respectively, and Fe 4 Mn 6 -Fc-MOFs achieved 92% removal. When the Fe/Mn rate continued to increase to 6:4, the removal efficiency decreased to 85%. It was noteworthy that the degradation efficiency of Mn-Fc-MOFs/PS reached 71%. These outcomes indicated that with the molar ratio of FeMn to a specific value, the catalytic performance would reach an ideal value on activating PS. Nevertheless, excessive Fe led to the inhibition for AO7 removal ). In addition, to better access the degradation ability of FeMn-Fc-MOFs, we calculated the reaction rate using a firstorder kinetic constant (Eqs. 1).
We also looked into the effects of catalyst dosage, PS concentration, and pH value. Figure 3b depicts the degradation of AO7 under various catalyst dosages. The degradation efficiency of AO7 was positively correlated with the dose of Fe 4 Mn 6 -Fc-MOFs in the range of 0 to 300 mg/L. Although the catalyst dose was raised from 400 to 500 mg/L, the degradation efficiency remained the same and the degradation rate was slightly decline. This was probably because the Fe 4 Mn 6 -Fc-MOFs catalyst clumped together, limiting the dispersion of reactants to the surface (Depeursinge et al. 2010). (1) The influence of PS concentration on the degradation of AO7 is shown in Fig. 3c. The degradation efficiency of AO7 tended to increase when PS concentration rose from 0 to 8 mmol/L. Additionally, AO7 degradation rates decreased with the PS concentrations increased to 15 mmol/L, because too much PS produced sulfate anions instead of SO 4 •− (Leng et al. 2014). Figure 3d suggests that the effect of solution pH on AO7 degradation efficiency. Fe 4 Mn 6 -Fc-MOFs was more positively for the removal of AO7 under acidic and neutral conditions than in alkaline ones which was in agreement with earlier research (Xiao et al. 2016). This might be because better sorption performance of Fe 4 Mn 6 -Fc-MOFs and the redox potential of SO 4 •− and • OH decreased when pH increased from 3.13 to 10.11, especially that of • OH. Additionally, the depletion of SO 4 •− occurred in alkaline circumstances as a result of the reaction between the -OH and SO 4 •− generated by activated persulfate (J. Wang et al. 2016). Therefore, alkaline conditions are harmful for the deterioration of AO7.

Mechanism of persulfate activation
The main active species generated during degradation were of great help in the elucidation of the reaction mechanism. To further clarify the main active radicals in Fe 4 Mn 6 -Fc-MOFs system and the mechanism of pollutants degradation, quenching experiments and EPR spectroscopy were carried out. Firstly, quenching experiments were performed to ascertain the main active species for the effective degradation of AO7. Generally, ethanol was used as selective probes for SO 4 •− and • OH radicals. TBA has a stronger affinity for • OH scavenging than SO 4 •− (Pi et al. 2018). Due to this, the difference in the degradation yield of AO7 caused by the addition of ethanol and TBA indicated the performance of sulfate. As displayed in Fig. 4a, both TBA and ethanol have a remarkable inhibitory effect on the degradation of AO7. When TBA and ethanol was added into Fe 4 Mn 6 -Fc-MOFs/ PS system, the removal rate decreased from 90 to 59% and 42%, respectively. Therefore, the addition of TBA and ethanol had great effects on AO7 removal, indicating that SO 4 •− and • OH radicals fulfilled a vital role in the degradation process.
Secondly, EPR was utilized to verify the free radical species produced by the Fe 4 Mn 6 -Fc-MOFs/PS system as it degraded AO7 (Dong et al. 2022). As shown in Fig. 4b, there was no peak in the test group of the PS. A sequence of peaks appeared after the addition of the Fe 4 Mn 6 -Fc-MOFs, indicating the generation of ROS. The DMPO-SO 4 •− signals with characteristic peaks 1:1:1:1 exhibited the presence of SO 4 •− . Obviously, the characteristic peaks 1:2:2:1 were the signals of DMPO-• OH. These indicated that the Fe 4 Mn 6 -Fc-MOFs/ PS system had indeed generated SO 4 •− and • OH. This outcome holds true for the quenching experiments.
Electrochemical testing is an effective method to characterize the catalytic activity of catalysts, especially the electron transfer ability (R. Wang et al. 2022). To further investigate the differences in catalytic activity of the prepared catalysts, electrochemical experiments were conducted. One of the most important properties of ferrocene and its derivatives was their powerful reducing ability, and the cyclic voltammetry (CV) experiment confirmed this. Figure 5 shows the CV curves of electrodes made of four catalysts in the voltage range − 0.8 to 1.0 V. The oxidation and reduction peaks changed in the positive and negative orientations. And a symmetric peak appeared, indicating that the redox reaction was reversible. Notably, the Fe 4 Mn 6 -Fc-MOFs exhibited a higher current than that of other catalysts. Meanwhile, the voltage difference in the curve of Fe 4 Mn 6 -Fc-MOFs was dramatically reduced, indicating that the conductivity of Fe 4 Mn 6 -Fc-MOFs is superior to the others. These results also suggested that Fe 4 Mn 6 -Fc-MOFs had better electron transfer capability, thereby promoting the activation of PS, which is consistent with the reported results (M. Li et al. 2009).
EIS measurements were used to analyze the capacitive response of the electrodes (X. Sun et al. 2011). The semicircle was substantially different in the high-frequency band, as seen in Fig. 5. The radius of Fe 4 Mn 6 -Fc-MOFs was the smallest in the samples, indicating that Fe 4 Mn 6 -Fc-MOFs had higher charge separation efficiency. The results showed that the catalytic activity was the best when the molar ratio of Fe to Mn was 4:6. Under the condition of the molar ratio 4:6, the effective and adequate interaction of Mn and Fe made the best of their synergism. And a similar outcome can be found in the CV test.
Next, to further account for the activation mechanism of Fe 4 Mn 6 -Fc-MOFs on PS, the chemical valence of active substances (i.e., Fe and Mn) before and after reaction were analyzed by XPS in the Fe 4 Mn 6 -Fc-MOFs/PS system. The full range spectrum illustrated the existence of C 1 s, O 1 s, Mn 2p, and Fe 2p, indicating that catalysts were stable after   (Fig. 6a). As exhibited in Fig. 6b, the convolution diagram of Fe 2p distribution before and after reaction Fe 4 Mn 6 -Fc-MOFs was four peaks at 705.7 eV,708.7 eV,719.4 eV,and 725.6 eV,corresponding to Fe(II) and Fe(III) with atomic ratios of 60.42:39.58% ahead of the reaction and 66.24:33.76% after the reaction (Yamashita and Hayes 2008). This illustrated that Fe(III) was converted to Fe(II) on the Fe 4 Mn 6 -Fc-MOFs' surface (up from 60.42 before to 66.24% after the reaction). These changes in elemental composition might be attributed to redox reactions between Fe(III) and PS on the sample surface. Similarly, as shown in Fig. 6c, Mn 2p 3/2 distributed at 638.2 eV, 640.9 eV, and 643.6 eV before and after reaction in Fe 4 Mn 6 -Fc-MOFs disintegrated into three Mn 2p 3/2 peaks (Iwanowski et al. 2004). These correspond to Mn(II), Mn(III), and Mn(IV) with atomic ratios of 47.22:25.00:27.77% before the reaction and 42.82:28.96:28.21% after the reaction. Following the reaction, the fraction of Mn(II) decreased significantly, from 47.22 to 42.82%. Correspondingly, the proportions of Mn(III) and Mn(IV) rose from 25.00 and 27.77 to 28.96% and 28.21%, respectively.
As a result of the aforementioned results and previous literature, a potential mechanism of FeMn-Fc-MOFs/PS was proposed and presented in Fig. 7. According to these results above, the activation of PS occurred on the surface of the catalyst. Low-valent metals lost the outermost electron, which was transferred to PS for the cleavage of the O-O bond to generate SO 4 •− . Correspondingly, Fe(III), Mn(III), and Mn(IV) are formed (Eqs. 2-4). The redox potential of PS (2.0 V) is much higher than that of Fe(III) (0.77 V), Mn(III) (1.51 V), and Mn(IV) (0.95 V), so the reduction of Fe(III), Mn(III), and Mn(IV) were thermodynamically unfavorable. The ferrocenyl (≡ Fc) in MOFs was a redox mediator with excellent performance, which accelerated the electron transfer through reversible redox reactions and led to the rapid reduction of Fe(III), Mn(III), and Mn(IV) (Eqs. 5-7) (Lv et al. 2022). The more reduced transition metal ions were produced, the more amount of SO 4 •− in the In addition, the ferrocenyl in the catalyst also activated PS to produce SO 4 •− through its redox cycle (Eqs. 10 and 11). In the above process, the generated SO 4 •− reacts with H 2 O/ OHto form • OH (Eqs. 12 and 13). And the generated • OH and SO 4 . •− decomposed AO7 to various intermediates by several reactions (involving electron transfer, electrophilic/ free addition). Eventually, the intermediates were mineralized to CO 2 and H 2 O (Eq. 14)

Stability and reusability of Fe 4 Mn 6 -Fc-MOFs
The recyclability and stability of Fe 4 Mn 6 -Fc-MOFs were evaluated. After catalytic degradation, the catalyst was centrifuged, washed several times with ethanol and distilled water, and lastly aired at 75 °C for recycling. As displayed in Fig. 8, fresh Fe 4 Mn 6 -Fc-MOFs showed excellent catalytic activity in PS activation, and more than 91% of AO7 was degraded within 60 min. For the second to fifth cycles, the degradation efficiency of AO7 is 87%, 84%, 82%, and 79%, respectively, which still had good degradation performance. In addition, the minor decrease in Fe 4 Mn 6 -Fc-MOF catalytic performance could be attributed to the block of active sites caused by degradative products deposited. For the stability of the catalyst, we conducted a systematic study. According to the results of the ICP-OES, the elements Fe and Mn were almost not dissolved. All in all, although the degradation efficiency had decreased  after five cycles of experiments, Fe 4 Mn 6 -Fc-MOFs still had a high overall degradation efficiency and can degrade 79% of AO7 in 60 min, demonstrating that Fe 4 Mn 6 -Fc-MOFs has excellent degradation ability and stability.

Conclusion
In conclusion, a series of FeMn-Fc-MOFs catalysts with different molar ratios were synthesized by a modified hydrothermal method. The prepared catalysts showed better catalytic activity, among which the catalytic properties of Fe 4 Mn 6 -Fc-MOFs were the most prominent due to the better electron transfer capacity, which made the catalytic activation center have better performance. The presence of Fc accelerated the redox cycle and the synergistic effect of Fe and Mn accelerated the reversible redox reaction in the degradation process. In addition, the catalyst had good stability and reusability. The above results showed that Fe 4 Mn 6 -Fc-MOFs had great degradation potential for environmental pollutants.