Excellent Performance of Fe78Si9B13 Metallic Glass for Activating Peroxymonosulfate in Degradation of Naphthol Green B

: The functional application of metallic glasses in the catalytic ﬁeld has widely attracted research attention due to its unique atomic structure compared to crystalline materials. It has been reported that metallic glasses can effectively activate H 2 O 2 and persulfate, yet the activation of peroxymonosulfate by metallic glasses is not studied well. In this work, the metallic glass with atomic composition of Fe 78 Si 9 B 13 was applied for investigating the peroxymonosulfate (PMS) activation on degradation of naphthol green B (NGB) dye. The change of surface morphology indicated the important role of oxide ﬁlms during the dye degradation. The effects and ﬁrst-order kinetics model of various reaction parameters were evaluated systematically, including PMS concentration, catalyst dosage, irradiation intensity, and dye concentration. The results showed that about 98% of the dye removal rate could be achieved only within 10 min under rational conditions. The reaction kinetics k of 0.1339 min − 1 without ribbons was sharply improved to 0.3140 min − 1 by adding 0.5 g/L ribbons, indicating the superior activation ability of Fe 78 Si 9 B 13 metallic glass. The recycling experiment revealed that the Fe 78 Si 9 B 13 ribbons exhibited the excellent surface stability and catalytic reusability for activating PMS even after reused for 10th run.


Introduction
In addition to application as structural materials, the metallic glasses, having their far-from-equilibrium nature and uniquely amorphous atomic packing structure, have been widely employed as functional materials in different fields, such as biodegradable implants [1], aerospace and automobile applications [2], and control devices [3]. Even thin film metallic glasses can be treated as a potential candidate for coatings [4][5][6]. Very recently, the metallic glasses have been demonstrated as a superior candidate in the field of wastewater treatment attributed to catalytically active and uniquely selective properties, in addition to the easy alteration of chemical compositions [7,8]. For example, ball-milled Co 78 Si 8 B 14 glassy powders exhibit excellent performance for completely degrading acid orange II in between 2 to 6 min under different conditions [9]; Mg 73 Zn 21.5 Ca 5.5 glassy powders are Naphthol green B (NGB), i.e., naphthol green, acid green 1, C.I. 10020 (C 30 H 15 FeN 3 Na 3 O 15 S 3 ), is an important acid dye. Owing to its bright color, water-fastness and easy coloration, NGB is widely applied as a coloring agent in the leather and fur industries and as a biological stain in modern industry. However, directly discharging NGB dye to the environment will cause serious pollution to aquatic life and mutagenicity to humans [43]. Due to the synthetic and complex aromatic structure, the NGB is difficult to be removed by biodegrading [44]. Furthermore, due to the high pH, high concentration, and high chroma, it is not economically feasible to remove the NGB from industrial wastewater effluents by using conventional methods [43]. Recently, some endeavors using AOPs have been employed to improve the removal efficiency of NGB, such as photocatalytic degradation of NGB by Al-doped ZnO [45], semiconducting Sb 2 S 3 [46] and Fe(III) induced by ligand-to-metal charge transfer excitation [47], and photo-Fenton degradation [48]. Nevertheless, those reported photodegradation of NGB cannot achieve satisfying efficiency (up to hours for complete degradation) while the photo-Fenton degradation using Fe 3+ produces large amounts of iron sludge.
To overcome the disadvantages of aforementioned NGB removal techniques, this work investigated the NGB dye degradation using Fe 78 Si 9 B 13 /PMS under UV-vis irradiation for promoting the degradation efficiency of NGB. Simultaneously, studying Fe-based metallic glass activation behavior towards peroxymonosulfate including the analysis for those important effects (catalyst dosage, PMS concentration, dye concentration, and irradiation intensity) can provide adequate understanding of heterogeneous sulfate-based oxidation process. In this work, the characterizations of the as-received and reused Fe 78 Si 9 B 13 ribbons were firstly studied by X-ray diffraction (XRD), ultraviolet visible diffuse reflectance spectrum (UV-vis DRS), scanning electron microscope (SEM), and energy-dispersive X-ray spectroscopy (EDS). The effects and corresponding pseudo-first-order kinetic model of catalyst dosage, PMS concentration, dye concentration, and irradiation intensity on the dye degradation efficiency were systematically examined and discussed in detail. In addition, the stability and reusability of the Fe 78 Si 9 B 13 ribbons were also studied.

Materials and Chemicals
The naphthol green B dye was supplied by Wenzhou Huaqiao Chemical Reagent Co., Ltd., Wenzhou, China. The main characteristics of the dye are summarized in Table 1. Oxone (KHSO 5 ·0.5KHSO 4 ·0.5K 2 SO 4 ) was purchased from Sigma-Aldrich. Ethanol (absolute) was obtained from Merck. In this work, the as-received Fe 78 Si 9 B 13 glassy ribbons, which were cut into 5 × 20 mm with a thickness of 30-40 µm, were manufactured by the melting spinning method. The master alloy ingot of Fe 78 Si 9 B 13 was firstly prepared by arc melting under a Ti-gettered argon atmosphere five times to achieve homogenous chemical composition. From the master alloy, the glassy ribbons were fabricated by ejecting melting master alloy on a single copper roller [49][50][51]. The wheel speed of the copper roller used was about 30 m/s. All the aqueous solutions throughout experiments were prepared using Milli-Q water (18.2 MΩ·cm). All the chemicals used in this work are in analytical grade and no further purification is required. with a thickness of 30-40 µm, were manufactured by the melting spinning method. The master alloy ingot of Fe78Si9B13 was firstly prepared by arc melting under a Ti-gettered argon atmosphere five times to achieve homogenous chemical composition. From the master alloy, the glassy ribbons were fabricated by ejecting melting master alloy on a single copper roller [49][50][51]. The wheel speed of the copper roller used was about 30 m/s. All the aqueous solutions throughout experiments were prepared using Milli-Q water (18.2 ΜΩ·cm). All the chemicals used in this work are in analytical grade and no further purification is required.

Methods
The structure of the as-received and reused Fe 78 Si 9 B 13 glassy ribbons were identified by XRD. The XRD data were collected on a PANalytical Empyrean diffractometer (Eindhoven, The Netherlands) with monochromated Co-Kα radiation (λ = 0.1789 nm) operated at 40 kV and 40 mA. The UV-vis DRS (Shelton, CT, USA) was employed to examine the surface characterization of as-received and after reused ribbons at the 200-800 nm range using BaSO 4 as the reference. The surface morphology of before and reused Fe 78 Si 9 B 13 ribbons were analyzed by SEM equipped with EDS (JEOL 6000, Tokyo, Japan) with an accelerating voltage of 15 kV. To analyze the catalytic stability and reusability, the Fe 78 Si 9 B 13 ribbon samples were reused 10 times. After each time, the reused ribbons were washed by Milli-Q water in an ultrasonic cleaner for 3 min, then further carefully cleaned by analytical grade alcohol. All the reused ribbons were preserved in the absolute ethanol for further characterization.
For the dye degradation experiments, 1000 ppm NGB dye solution was firstly prepared and stored in a 500 mL flask for preservation. Then the specific volume of NGB dye solution of 1000 ppm was extracted by a macropipette (SocorexIsba S.A., Ecublens, Switzerland) and diluted to a required dye concentration (10 ppm, 20 ppm, 50 ppm, and 100 ppm). During all the experiments, ribbons with a specific weight (0.03 g, 0.05 g, 0.1 g, and 0.2 g) were immersed into a 100 mL NGB solution under continuous mechanical stirring by Vortex-Genie 2 mixer (Scientific Industries, Inc., New York, NY, USA). The irradiation was provided by a 300 W Xeon simulated solar light lamp (Perfectlight Scientific Pty Ltd., Beijing, China). The pH value of dye solution throughout this work was about 5.8, measured by an Oakton pH/conductivity meter (PC 2700, Cole-Parmer, Vernon Hills, IL, USA). Approximate 3.5 mL NGB dye sample was taken from the beaker at the predetermined time intervals (i.e., 1,2,4,6,8,10,12,15,20, and 30 min), then injected into a 3.5 mL quartz cuvette with a pathlength 10 mm, following scanning in turn using a Perkin Elmer Lambda 35 UV-vis spectrometer (Shelton, CT, USA). The wavelength of light absorbance λ max of NGB was measured at the 408 nm. After the addition of PMS, the peak shifts from 714 to 408 nm due to the acidic nature of PMS. Dye degradation efficiency was calculated by Equation (1): where C 0 and C are the initial concentration and the concentration at time t (min) of NGB, respectively. The NGB decomposition was fitted by the pseudo-first-order kinetic model as shown in Equation (2), similar to other reports [26,52]: where k obs is the calculated reaction rate. The equation of the kinetic model can be rewritten as Equation (3):

Characterizations of Fe 78 Si 9 B 13
The Fe 78 Si 9 B 13 ribbon samples were reused 10 times for degrading NGB under the same conditions: PMS concentration of 1.0 mmol/L; ribbons dosage of 0.5 g/L; irradiation intensity of 7.7 µW/cm 2 ; and dye concentration of 50 ppm. Then the ribbons were characterized at the specific reused time. Figure 1a shows the XRD patterns of as-received, fifth, and tenth run reused Fe 78 Si 9 B 13 ribbons. As can be seen, all the XRD patterns present a broad diffraction peak at a range of 2θ = 40-60 • , indicating that they are mainly composed of the amorphous phase [53][54][55][56][57]. Clearly, the maximum diffraction peaks at 2θ = 52.6 • and 95.7 • exhibit a higher diffraction intensity with the increased reused times, which is because crystallized α-Fe is gradually formed on the surface of the Fe 78 Si 9 B 13 ribbons during NGB degradation process under the UV-vis irradiation [58,59]. Figure 1b presents the UV-vis DRS patterns of the as-received, fifth and tenth run reused Fe 78 Si 9 B 13 ribbons. It is known that the tetrahedral or octahedral isolated iron species will cause the bands between 200 and 300 nm, while the octahedral irons in the form of Fe 3+ x O y complexes result in the bands between 300 and 450 nm [60,61]. In addition, large iron oxide aggregates may lead to the bands greater than 450 nm [60,61]. In this work, the peaks between 200 and 350 nm of the as-received Fe 78 Si 9 B 13 ribbons indicate isolated iron species are distributed homogeneously, while no obvious peak after 450 nm means no iron oxide aggregates on the surface. After reused fifth and tenth run, slight peaks after 550 nm are observed in these two Fe 78 Si 9 B 13 ribbons, indicating that partly iron oxides aggregates (excepting α-Fe) are gradually formed on both ribbons surface. Compared with the as-received Fe 78 Si 9 B 13 ribbons, the obvious peaks between 200 and 270 nm of the fifth and tenth run reused ribbons indicate that the isolated iron species still distribute homogeneously on the surface, and peaks between 300 and 360 nm are due to octahedral irons presenting as Fe 3+ x O y complexes form. the tetrahedral or octahedral isolated iron species will cause the bands between 200 and 300 nm, while the octahedral irons in the form of Fe 3+ xOy complexes result in the bands between 300 and 450 nm [60,61]. In addition, large iron oxide aggregates may lead to the bands greater than 450 nm [60,61]. In this work, the peaks between 200 and 350 nm of the as-received Fe78Si9B13 ribbons indicate isolated iron species are distributed homogeneously, while no obvious peak after 450 nm means no iron oxide aggregates on the surface. After reused fifth and tenth run, slight peaks after 550 nm are observed in these two Fe78Si9B13 ribbons, indicating that partly iron oxides aggregates (excepting α-Fe) are gradually formed on both ribbons surface. Compared with the as-received Fe78Si9B13 ribbons, the obvious peaks between 200 and 270 nm of the fifth and tenth run reused ribbons indicate that the isolated iron species still distribute homogeneously on the surface, and peaks between 300 and 360 nm are due to octahedral irons presenting as Fe 3+ xOy complexes form.   Figure 2a, the free surface of the as-received ribbons is practically smooth without any apparent defects. Figure 2b shows the corresponding EDS analysis of as-received ribbons. Apparently, the peaks in the EDS are mainly formed by the element of Fe and Si, and Fe occupies the dominant composition, which is similar to the nominal composition of Fe78Si9B13 ribbons (domination of Fe). After reused fifth run, the ribbon surface in Figure 2c presents a clear boundary and different layers can be distinguished. The bright area seems like the films which is not adherent to the ribbon surface. According to Figure 2d, it is obvious that after reused fifth run, the ribbon surface has a high concentration of O compared with Figure 2b. Based on the previous research [23,24], it can be  As seen from Figure 2a, the free surface of the as-received ribbons is practically smooth without any apparent defects. Figure 2b shows the corresponding EDS analysis of as-received ribbons. Apparently, the peaks in the EDS are mainly formed by the element of Fe and Si, and Fe occupies the dominant composition, which is similar to the nominal composition of Fe 78 Si 9 B 13 ribbons (domination of Fe). After reused fifth run, the ribbon surface in Figure 2c presents a clear boundary and different layers can be distinguished. The bright area seems like the films which is not adherent to the ribbon surface. According to Figure 2d, it is obvious that after reused fifth run, the ribbon surface has a high concentration of O compared with Figure 2b. Based on the previous research [23,24], it can be confirmed that the region next to the boundary connecting the bright area is covered by the oxide films (iron oxides and silicon dioxide). With the reused time increasing to tenth run, comparing Figure 2e with Figure 2c, the previously distinct boundary disappears. Instead, there are many small pits that can be observed on the ribbon surface (as the enlarged surface in the Figure 2e). Although, the oxide films in the Figure 2e cannot be distinguished easily from its EDS result (Figure 2f), the relatively high concentration of O still demonstrates that the tenth run reused surface is covered by oxide films, but the structure of the film has changed. According to the transformation of surface morphology from as-received ribbons to tenth run reused ribbons, the mechanism of surface change can be concluded as: (1) the as-received ribbon surface is very smooth without any obvious defects; (2) with the reused time increasing, the oxide films due to a series of reactions between Fe and Si, and O 2 are gradually formed on the reused ribbons surface. These oxide films easily fall off from the ribbons surface during the degradation process when being stirred by the mixer; (3) with the loose films exuviating, fresh Fe is exposed to air. Then the Fe continuously goes through a series of reactions to form new compact films; (4) the corrosion of Fe becomes severer with further reactions leading to the formation of pits although the new compact film has strong ability to protect the buried Fe. From Figure 2d,f, compared with the as-received ribbons, the intensity of O increases obviously leading to the total intensity of peaks of Fe and Si on the fifth or tenth run reused ribbons surface is sharply reduced. The possible reasons are: (1) the isolated Fe at the as-received ribbons is gradually corroded during degradation process, leading to the formation of dissociative iron species (Fe 2+ and Fe 3+ ) and therefore the detected Fe atom on the surface decreases; (2) the application of UV-vis irradiation assists in transferring electron on the 4S 2 orbital of ribbons due to the extra energy input [24]; (3) PMS has the ability to directly react with isolated Fe. Those corrosion behaviors occur on the ribbons surface concurrently, resulting in an effective and sufficient supplement of iron source to activate PMS. confirmed that the region next to the boundary connecting the bright area is covered by the oxide films (iron oxides and silicon dioxide). With the reused time increasing to tenth run, comparing Figure 2e with Figure 2c, the previously distinct boundary disappears. Instead, there are many small pits that can be observed on the ribbon surface (as the enlarged surface in the Figure 2e). Although, the oxide films in the Figure 2e cannot be distinguished easily from its EDS result (Figure 2f), the relatively high concentration of O still demonstrates that the tenth run reused surface is covered by oxide films, but the structure of the film has changed. According to the transformation of surface morphology from as-received ribbons to tenth run reused ribbons, the mechanism of surface change can be concluded as: (1) the as-received ribbon surface is very smooth without any obvious defects; (2) with the reused time increasing, the oxide films due to a series of reactions between Fe and Si, and O2 are gradually formed on the reused ribbons surface. These oxide films easily fall off from the ribbons surface during the degradation process when being stirred by the mixer; (3) with the loose films exuviating, fresh Fe is exposed to air. Then the Fe continuously goes through a series of reactions to form new compact films; (4) the corrosion of Fe becomes severer with further reactions leading to the formation of pits although the new compact film has strong ability to protect the buried Fe. From Figure 2d,f, compared with the as-received ribbons, the intensity of O increases obviously leading to the total intensity of peaks of Fe and Si on the fifth or tenth run reused ribbons surface is sharply reduced. The possible reasons are: (1) the isolated Fe at the as-received ribbons is gradually corroded during degradation process, leading to the formation of dissociative iron species (Fe 2+ and Fe 3+ ) and therefore the detected Fe atom on the surface decreases; (2) the application of UV-vis irradiation assists in transferring electron on the 4S 2 orbital of ribbons due to the extra energy input [24]; (3) PMS has the ability to directly react with isolated Fe. Those corrosion behaviors occur on the ribbons surface concurrently, resulting in an effective and sufficient supplement of iron source to activate PMS.   Figure 3 shows the UV-vis spectra of NGB degradation at 0.5 g/L Fe 78 Si 9 B 13 ribbons within different time intervals (PMS: 1.0 mmol/L, irradiation intensity: 7.7 µW/cm 2 , dye concentration: 50 ppm). Apparently, with the addition of PMS, a large blue-shift of maximum absorbance peak occurs immediately, from 714 to 408 nm. From the Figure 3, it can be seen that the maximum peak at λ = 408 nm decreases very fast and become invisible after 10 min reaction time, indicating a progressive removal of the chromophore (-NO) in NGB molecules and a completion after 10 min [47]. In addition, no obvious peak is formed after 10 min and all the previous peaks disappear in both ultraviolet and visible light regions, suggesting that benzene ring opening and cleavage of the central carbon happen at the same time. The result in the UV-vis spectra reveals that the reaction between Fe 78 Si 9 B 13 ribbons and PMS under UV-vis irradiation can effectively eliminate NGB dye within a very short time, converting dye molecules to harmless inorganic substances.  Figure 3 shows the UV-vis spectra of NGB degradation at 0.5 g/L Fe78Si9B13 ribbons within different time intervals (PMS: 1.0 mmol/L, irradiation intensity: 7.7 µW/cm 2 , dye concentration: 50 ppm). Apparently, with the addition of PMS, a large blue-shift of maximum absorbance peak occurs immediately, from 714 to 408 nm. From the Figure 3, it can be seen that the maximum peak at λ = 408 nm decreases very fast and become invisible after 10 min reaction time, indicating a progressive removal of the chromophore (-NO) in NGB molecules and a completion after 10 min [47]. In addition, no obvious peak is formed after 10 min and all the previous peaks disappear in both ultraviolet and visible light regions, suggesting that benzene ring opening and cleavage of the central carbon happen at the same time. The result in the UV-vis spectra reveals that the reaction between Fe78Si9B13 ribbons and PMS under UV-vis irradiation can effectively eliminate NGB dye within a very short time, converting dye molecules to harmless inorganic substances. According to the result in the Figure 3, the removal efficiency of NGB in this work is compared with the other results, which can be seen from the Table 2. It is clear that NGB dye removal using Fe78Si9B13/PMS under UV-vis irradiation in this work has a faster efficiency which achieves full decolorization within 10 min with a reaction rate kobs = 3.14 × 10 −1 min −1 , while the other results receive a very low efficiency. The required time of fully dye removal is double and more than sextuple for the reported photo-Fenton process [48] and photocatalysis [45,46], respectively, than this work, although all of these methods are based on AOPs. Obviously, the faster NGB degradation by Fe78Si9B13/PMS system is attributed to aggressively attack by large quantities of reactive species. The comparable result further indicates that Fe78Si9B13 metallic glass is a superior catalyst for activating PMS under UV-vis irradiation.  According to the result in the Figure 3, the removal efficiency of NGB in this work is compared with the other results, which can be seen from the Table 2. It is clear that NGB dye removal using Fe 78 Si 9 B 13 /PMS under UV-vis irradiation in this work has a faster efficiency which achieves full decolorization within 10 min with a reaction rate k obs = 3.14 × 10 −1 min −1 , while the other results receive a very low efficiency. The required time of fully dye removal is double and more than sextuple for the reported photo-Fenton process [48] and photocatalysis [45,46], respectively, than this work, although all of these methods are based on AOPs. Obviously, the faster NGB degradation by Fe 78 Si 9 B 13 /PMS system is attributed to aggressively attack by large quantities of reactive species. The comparable result further indicates that Fe 78 Si 9 B 13 metallic glass is a superior catalyst for activating PMS under UV-vis irradiation.

Single Parameter Effect on Dye Degradation
The degradation process is strongly influenced by the following main operating parameters: PMS concentration; catalyst dosage; irradiation intensity; and dye concentration. In order to study the effect of these parameters systematically, it was investigated by altering the specific parameter while other parameters were constant. All the investigations were compared with the condition (PMS concentration: 1.0 mmol/L; Fe 78 Si 9 B 13 ribbons dosage: 0.5 g/L; UV-vis irradiation intensity: 7.7 µW/cm 2 ; dye concentration: 50 ppm).

Effect of PMS Concentration
Among the specific reaction parameters of the dye degradation process, the initial concentration of PMS is a key parameter. Figure 4 shows the compared results of PMS concentrations vary from 0 to 2.0 mmol/L. It is found from Figure 4a that only less than 10% of color removal rate could be reached within 30 min when solely using Fe 78 Si 9 B 13 ribbons at the dosage of 0.5 g/L. It is postulated that the color removal is the consequence of the adsorption property of the ribbons [12], in addition to poor NGB self-photosensitization under UV-vis irradiation [62]. The color removal rate increases remarkably with adding PMS from 0 to 0.25 mmol/L. However, with the low concentration of PMS (0.25 mmol/L), the removal rate of NGB only achieves less than 80% within 15 min and no further increase occurs along the reaction time, while the degradation rate achieves about 94% within 8 min at the 2.0 mmol/L PMS. The first-order kinetic model is shown in Figure 4b and the corresponding data are summarized in Table 3. It can be found that the reaction kinetics (k) value of 1.0 mmol/L PMS is 0.3140 min −1 , which is twice as high as the value of PMS (0.25 mmol/L), while the k value is just slightly improved to 0.

Single Parameter Effect on Dye Degradation
The degradation process is strongly influenced by the following main operating parameters: PMS concentration; catalyst dosage; irradiation intensity; and dye concentration. In order to study the effect of these parameters systematically, it was investigated by altering the specific parameter while other parameters were constant. All the investigations were compared with the condition (PMS concentration: 1.0 mmol/L; Fe78Si9B13 ribbons dosage: 0.5 g/L; UV-vis irradiation intensity: 7.7 µW/cm 2 ; dye concentration: 50 ppm).

Effect of PMS Concentration
Among the specific reaction parameters of the dye degradation process, the initial concentration of PMS is a key parameter. Figure 4 shows the compared results of PMS concentrations vary from 0 to 2.0 mmol/L. It is found from Figure 4a that only less than 10% of color removal rate could be reached within 30 min when solely using Fe78Si9B13 ribbons at the dosage of 0.5 g/L. It is postulated that the color removal is the consequence of the adsorption property of the ribbons [12], in addition to poor NGB self-photosensitization under UV-vis irradiation [62]. The color removal rate increases remarkably with adding PMS from 0 to 0.25 mmol/L. However, with the low concentration of PMS (0.25 mmol/L), the removal rate of NGB only achieves less than 80% within 15 min and no further increase occurs along the reaction time, while the degradation rate achieves about 94% within 8 min at the 2.0 mmol/L PMS. The first-order kinetic model is shown in Figure 4b and the corresponding data are summarized in Table 3. It can be found that the reaction kinetics (k) value of 1.0 mmol/L PMS is 0.3140 min −1 , which is twice as high as the value of PMS (0.25 mmol/L), while the k value is just slightly improved to 0.     Figure 5 shows the effect of Fe 78 Si 9 B 13 dosage ranging from 0 to 2.0 g/L on the color removal of NGB dye under the following experimental conditions (PMS concentration: 1.0 mmol/L; irradiation intensity: 7.7 µW/cm 2 ; and dye concentration: 50 ppm). It can be seen from Figure 5a, the NGB dye is almost completely degraded within 10 min at the ribbons dosage of 0.5 g/L, whereas 30 min is required to complete the degradation without Fe 78 Si 9 B 13 ribbons, indicating that the Fe 78 Si 9 B 13 ribbons can effectively activate PMS to produce reactive species. Further increases of ribbon dosage leads to slight effect of complete reaction time, where all of the reactions with addition of ribbons can reach more than 94% of color removal rate within 10 min. The first-order kinetic model is shown in Figure 5b and the corresponding data are summarized in Table 4. It can be seen from Table 4 that the reaction kinetics (k) value without Fe 78 Si 9 B 13 is 0.1489 min −1 while it is the double with adding ribbons dosage of 0.5 g/L, also confirming that PMS can be well activated by the Fe 78 Si 9 B 13 metallic glass. It is known that Fe 0 in the ribbons can directly react with PMS to produce SO 4 • − in the PMS solution (Equation (5)) [63]. Furthermore, with the enhancement of UV-vis irradiation, the corrosion of Fe 78 Si 9 B 13 ribbons are accelerated to provide more available Fe 2+ (Equation (6)) [24]. The produced Fe 2+ further reacts with PMS (Equation (7)). Sustainable generation of Fe 2+ can be also achieved by Fe 3+ reacting with Fe 0 (Equation (8)). As a result, the degradation rate of NGB is highly promoted.   Figure 5 shows the effect of Fe78Si9B13 dosage ranging from 0 to 2.0 g/L on the color removal of NGB dye under the following experimental conditions (PMS concentration: 1.0 mmol/L; irradiation intensity: 7.7 µW/cm 2 ; and dye concentration: 50 ppm). It can be seen from Figure 5a, the NGB dye is almost completely degraded within 10 min at the ribbons dosage of 0.5 g/L, whereas 30 min is required to complete the degradation without Fe78Si9B13 ribbons, indicating that the Fe78Si9B13 ribbons can effectively activate PMS to produce reactive species. Further increases of ribbon dosage leads to slight effect of complete reaction time, where all of the reactions with addition of ribbons can reach more than 94% of color removal rate within 10 min. The first-order kinetic model is shown in Figure  5b and the corresponding data are summarized in Table 4. It can be seen from Table 4 that the reaction kinetics (k) value without Fe78Si9B13 is 0.1489 min −1 while it is the double with adding ribbons dosage of 0.5g/L, also confirming that PMS can be well activated by the Fe78Si9B13 metallic glass. It is known that Fe 0 in the ribbons can directly react with PMS to produce SO4• − in the PMS solution (Equation (5)) [63]. Furthermore, with the enhancement of UV-vis irradiation, the corrosion of Fe78Si9B13 ribbons are accelerated to provide more available Fe 2+ (Equation (6)) [24]. The produced Fe 2+ further reacts with PMS (Equation (7)). Sustainable generation of Fe 2+ can be also achieved by Fe 3+ reacting with Fe 0 (Equation (8)). As a result, the degradation rate of NGB is highly promoted.
Fe 0 + 2Fe 3+ → 3Fe 2+ (8)  Figure 6a shows the effect of irradiation intensity varies from 0 to 14.8 µW/cm 2 on color removal and the degradation rates of NGB dye using PMS activation by Fe 78 Si 9 B 13 ribbons. The first-order kinetic model is shown in Figure 6b and the corresponding data are summarized in Table 5. It is observed from Figure 6a that the dye removal rate with Fe 78 Si 9 B 13 ribbons alone present a relative fast efficiency, indicating that Fe 78 Si 9 B 13 metallic glass can activate PMS in a fast rate. On the other hand, the color removal rate reaches about 98% when the irradiation intensity increases from 0 to 7.7 µW/cm 2 , revealing that UV-vis irradiation can highly enhance the reactions. It is interesting to find that further increasing the irradiation intensity from 7.7 to 14.8 µW/cm 2 results in a rapid increase of color removal rate at the first 4 min, especially at 14.8 µW/cm 2 , but all of them can almost complete the reaction in 8 min. As seen from Table 5, the color removal is more than 90% at 8 min with the irradiation intensity of 7.7 µW/cm 2 or higher. Compared with k = 0.5247 min −1 using 14.8 µW/cm 2 irradiation intensity, the reaction kinetics (k) is 0.0816 min −1 without the irradiation. It can be confirmed that the irradiation can effectively promote the color removal rate. The reasons for the enhancement by application of UV-vis irradiation are: (1) direct activation of PMS by UV-vis irradiation to produce SO 4 • − and •OH (Equation (9)); (2) sustainable generation of Fe 2+ from Fe 3+ under UV-vis irradiation and •OH at the same time (Equation (10)) [24]; (3) reaction between PMS and produced Fe 2+ (Equation (7)).

Effect of Irradiation Intensity
Fe(OH) 2+ + hv → •OH + Fe 2+ (10)   Figure 7a shows the color removal rate of the NGB dye with the concentration ranging from 10 to 100 ppm when other conditions are constant (PMS concentration: 1.0 mmol/L, Fe78Si9B13 dosage: 0.5 g/L and irradiation intensity: 7.7 µW/cm 2 ). It is clear that the color removal rate decreases with the increase of NGB concentration. Compared with about 95% color removal rate at 6 min when degrading 10 ppm NGB, there is less than 70% color removal rate of 100 ppm NGB can be achieved at the same time. All the decolorization processes finish and reach the steady state within 12 min. Figure 7b shows the first-order kinetic model with the different dye concentrations and Table 6 shows the corresponding rate constants k. As seen from Table 6 the reaction kinetics k drops sharply from the highest value k = 0.4566 min −1 at 10 ppm dye concentration to k = 0.2030 min −1 at 100 ppm dye concentration. It means dye effluents can be decomposed better at lower concentration. The possible reason is higher dye concentration may block the light transmission to PMS, which results in lower activation of PMS to generate enough radicals [26]. Therefore, higher dye concentration has a lower color removal rate.   Figure 7a shows the color removal rate of the NGB dye with the concentration ranging from 10 to 100 ppm when other conditions are constant (PMS concentration: 1.0 mmol/L, Fe 78 Si 9 B 13 dosage: 0.5 g/L and irradiation intensity: 7.7 µW/cm 2 ). It is clear that the color removal rate decreases with the increase of NGB concentration. Compared with about 95% color removal rate at 6 min when degrading 10 ppm NGB, there is less than 70% color removal rate of 100 ppm NGB can be achieved at the same time. All the decolorization processes finish and reach the steady state within 12 min. Figure 7b shows the first-order kinetic model with the different dye concentrations and Table 6 shows the corresponding rate constants k. As seen from Table 6 the reaction kinetics k drops sharply from the highest value k = 0.4566 min −1 at 10 ppm dye concentration to k = 0.2030 min −1 at 100 ppm dye concentration. It means dye effluents can be decomposed better at lower concentration. The possible reason is higher dye concentration may block the light transmission to PMS, which results in lower activation of PMS to generate enough radicals [26]. Therefore, higher dye concentration has a lower color removal rate.

Stability and Reusability
In order to study the surface stability and catalytic reusability of Fe78Si9B13 ribbons, the same ribbons are reused 10 times. All the recycling experiments in this work are carried out on the same conditions (PMS concentration: 1.0 mmol/L; Fe78Si9B13 dosage: 0.5 g/L; irradiation intensity: 7.7 µW/cm 2 ; dye concentration: 50 ppm). Figure 8 shows the results of first, third, fifth, and tenth run of dye degradation with Fe78Si9B13 ribbons. The corresponding data of the first-order kinetic model are summarized in Table 7. It can be seen that the NGB degradation rate is about 92% within 8 min for the first run and only very slightly decreases during third, fifth, and tenth run. Comparing the color removal rate and the reaction kinetics (k) of the first run with the tenth run, the color removal rate slight drops from 92.1% to 89.4% at 8 min and the reaction rate k decreases from 0.2997 to 0.2571 min −1 . It is can be postulated that the product on the catalytic ribbons surface during the NGB degradation, as seen from Figure 2, which are crystallized α-Fe, iron oxides and SiO2, slightly decease the Fe78Si9B13 catalytic efficiency [12]. Although, the weak bonding films form firstly during the dye degradation, with the help of vortex-stirrer, it is easy to fall off from the catalytic ribbon surface, which make it possible to continuously provide sufficient iron source for activating PMS. It is easy to find that there is no obvious difference between the results of fifth and tenth run. It reveals that Fe78Si9B13 has excellent reusability in NGB degradation and can be run at least 10 times.

Stability and Reusability
In order to study the surface stability and catalytic reusability of Fe 78 Si 9 B 13 ribbons, the same ribbons are reused 10 times. All the recycling experiments in this work are carried out on the same conditions (PMS concentration: 1.0 mmol/L; Fe 78 Si 9 B 13 dosage: 0.5 g/L; irradiation intensity: 7.7 µW/cm 2 ; dye concentration: 50 ppm). Figure 8 shows the results of first, third, fifth, and tenth run of dye degradation with Fe 78 Si 9 B 13 ribbons. The corresponding data of the first-order kinetic model are summarized in Table 7. It can be seen that the NGB degradation rate is about 92% within 8 min for the first run and only very slightly decreases during third, fifth, and tenth run. Comparing the color removal rate and the reaction kinetics (k) of the first run with the tenth run, the color removal rate slight drops from 92.1% to 89.4% at 8 min and the reaction rate k decreases from 0.2997 to 0.2571 min −1 . It is can be postulated that the product on the catalytic ribbons surface during the NGB degradation, as seen from Figure 2, which are crystallized α-Fe, iron oxides and SiO 2 , slightly decease the Fe 78 Si 9 B 13 catalytic efficiency [12]. Although, the weak bonding films form firstly during the dye degradation, with the help of vortex-stirrer, it is easy to fall off from the catalytic ribbon surface, which make it possible to continuously provide sufficient iron source for activating PMS. It is easy to find that there is no obvious difference between the results of fifth and tenth run. It reveals that Fe 78 Si 9 B 13 has excellent reusability in NGB degradation and can be run at least 10 times.

Conclusions
In this work, it is indicated that PMS could be effectively activated by Fe78Si9B13 metallic glass for degrading NGB dye. The economic and efficient conditions can be concluded as 1.0 mmol/L PMS concentration, 0.5 g/L Fe78Si9B13 ribbons dosage, and 7.7 µW/cm 2 irradiation intensity. All the NGB degradation processes in this work fit well with the pseudo-first-order kinetic model. In addition, it is found that the NGB dye with a concentration of 50 ppm could be completely degraded within 10 min under the reasonable conditions of PMS concentration of 1.0 mmol/L, ribbons dosage of 0.5 g/L and irradiation intensity at 7.7 µW/cm 2 , which is faster than other reported materials [44,45].
According to the characterizations by XRD, UV-vis DRS, SEM, and EDS, the tenth run reused Fe78Si9B13 ribbons only present a slight decay, indicating that the metallic glass can well retain the amorphous nature although it has been used for activating PMS 10 times. The precipitated substances on the tenth run ribbons surface are confirmed as α-Fe, iron oxides, and silicon oxide. The change of ribbon surface morphology indicates that oxide films with weak bonding on the surface were formed first, following with falling off, formation of compact films, and pitting corrosion when increasing reused time of ribbons.
The color removal rate of NGB dye increases with the increase of PMS concentration, Fe78Si9B13 ribbons dosage and irradiation intensity. The recycling experiments show that Fe78Si9B13 metallic glass can be reused for activating PMS more than 10 times with the excellent dye degradation rate, presenting a superior surface stability and reusability. The Fe78Si9B13 metallic glass in this work receives a promising ability for activating PMS in NGB degradation and presents the potential wastewater treatment capacity for industrial scale in the future.

Conclusions
In this work, it is indicated that PMS could be effectively activated by Fe 78 Si 9 B 13 metallic glass for degrading NGB dye. The economic and efficient conditions can be concluded as 1.0 mmol/L PMS concentration, 0.5 g/L Fe 78 Si 9 B 13 ribbons dosage, and 7.7 µW/cm 2 irradiation intensity. All the NGB degradation processes in this work fit well with the pseudo-first-order kinetic model. In addition, it is found that the NGB dye with a concentration of 50 ppm could be completely degraded within 10 min under the reasonable conditions of PMS concentration of 1.0 mmol/L, ribbons dosage of 0.5 g/L and irradiation intensity at 7.7 µW/cm 2 , which is faster than other reported materials [44,45].
According to the characterizations by XRD, UV-vis DRS, SEM, and EDS, the tenth run reused Fe 78 Si 9 B 13 ribbons only present a slight decay, indicating that the metallic glass can well retain the amorphous nature although it has been used for activating PMS 10 times. The precipitated substances on the tenth run ribbons surface are confirmed as α-Fe, iron oxides, and silicon oxide. The change of ribbon surface morphology indicates that oxide films with weak bonding on the surface were formed first, following with falling off, formation of compact films, and pitting corrosion when increasing reused time of ribbons.
The color removal rate of NGB dye increases with the increase of PMS concentration, Fe 78 Si 9 B 13 ribbons dosage and irradiation intensity. The recycling experiments show that Fe 78 Si 9 B 13 metallic glass can be reused for activating PMS more than 10 times with the excellent dye degradation rate, presenting a superior surface stability and reusability. The Fe 78 Si 9 B 13 metallic glass in this work receives a promising ability for activating PMS in NGB degradation and presents the potential wastewater treatment capacity for industrial scale in the future.