Magnetic Activated-ATP@Fe3O4 Nanocomposite as an Efficient Fenton-Like Heterogeneous Catalyst for Degradation of Ethidium Bromide

Magnetic attapulgite-Fe3O4 nanocomposites (ATP-Fe3O4) were prepared by coprecipitation of Fe3O4 on ATP. The composites were characterized by scanning electron microscopey, X-ray diffractometry, Brunauer-Emmett-Teller analysis, X-ray photoelectron spectroscopy, energy dispersive spectrometer and transmission electron microscopy. Surface characterization showed that Fe3O4 particles with an average size of approximately 15 nm were successfully embedded in matrix of ATP. The capacity of the Fe3O4-activated ATP (A-ATP@Fe3O4) composites for catalytic degradation of ethidium bromide (EtBr, 80 mg/L) at different pH values, hydrogen peroxide (H2O2) concentrations, temperatures, and catalyst dosages was investigated. EtBr degradation kinetics studies indicated that the pseudo-first-order kinetic constant was 2.445 min−1 at T = 323 K and pH 2.0 with 30 mM H2O2, and 1.5 g/L of A-ATP@Fe3O4. Moreover, a regeneration study suggested that A-ATP@Fe3O4 maintained over 80% of its maximal EtBr degradation ability after five successive cycles. The effects of the iron concentrations and free radical scavengers on EtBr degradation were studied to reveal possible catalytic mechanisms of the A-ATP@Fe3O4 nanocomposites. Electron Paramagnetic Resonance revealed both hydroxyl (∙OH) and superoxide anion (∙O2−) radicals were involved in EtBr degradation. Radical scavenging experiment suggested EtBr degradation was mainly ascribed to ∙OH radicals, which was generated by reaction between Fe2+ and H2O2 on the surface of A-ATP@Fe3O4.

(P-ATP) and activated ATP (A-ATP) ( Fig. 1(a)) at 2θ = 13.7° and 19.8° are consistent with (200) and (040) planes of ATP 29 . The peaks at 2θ = 20.9° and 26.6° correspond to quartz (SiO 2 ) in P-ATP and A-ATP 28 30,31 . The intensities of the ATP peak at 2θ = 19.8° and the quartz peaks at 2θ = 20.9° and 26.6° were weaker for the Fe 3 O 4 -purified ATP (P-ATP@Fe 3 O 4 ) and Fe 3 O 4 -activated ATP (A-ATP@Fe 3 O 4 ), indicating that Fe 3 O 4 nano-particles were embedded in the ATP. But the characteristic reflections for ATP were observed in all of ATP@Fe 3 O 4 composites, suggesting that modification process did not destroy the characteristic structure of ATP.
The average crystallite size (D) of Fe 3 O 4 particles on the surface of A-ATP was 12.8 nm, as estimated with Scherrer equation 30 . The scanning electron microscopy (SEM) images ( Fig. 1(b,IV)) show that Fe 3 O 4 particles were distributed regularly on the rod-like structure of ATP. The SEM images indicate that the average size of the Fe 3 O 4 particles was about 15 nm, which matches well with XRD results.
Transmission electron microscopy (TEM) images show that the materials exhibited rod like structure with highly uniform rotundness and void sizes ( Fig. 2(a)), in agreement with the results of SEM analysis. The particles size ranged from 10-15 nm, which is consistent with the XRD result. Elemental mapping images (Fig. 2(c-g)) of A-ATP@Fe 3 O 4 confirmed that the presence of Fe and O atoms in nanocomposites with a content of above 30% and 50% ( Table 1). All of these results demonstrated that the Fe 3 O 4 nanoparticles were loaded on the ATP.
The nitrogen adsorption/desorption isotherms of Fe 3 O 4 , P-ATP@Fe 3 O 4 , and A-ATP@Fe 3 O 4 (Supplementary information Fig. S1(a)) all illustrated a typical type IV pattern with a bend of volume adsorption of nitrogen at a P/P 0 value of approximately 0.5 with a H 3 -type hysteresis loop. This pattern indicates the presence of mesoporous structure. In addition, the presence of mesoporous structure is also confirmed by the Barrett-Joyner-Halenda (BJH) corresponding pore size distribution curve (see Supplementary Fig. S1(b)). Furthermore, the typical type IV pattern with a H 3 -type hysteresis loop also illustrated that the nanomaterials comprised of aggregates (loose assemblages) of platelike (rod-like) particles forming slitlike pores 32 . The Brunauer-Emmett-Teller (BET) surface area, pore size, and pore volume of A-ATP@Fe 3 O 4 were 125.2745 m 2 /g, 11.80 nm, and 0.3695 cm 3 /g, respectively ( Table 2). In particular, the specific surface area was about 1.5 times larger than that of other reported catalysts 33 .
Four peaks were observed in the Fe 2p spectrum of A-ATP@Fe 3 O 4 ( Fig. 3(b)). The Fe 2p 3/2 peak with a BE of 710.7 eV was indicative of Fe 3+ octahedral species 34 . The BE of 724.8 eV for Fe 2p 1/2 indicated the presence of octahedrally coordinated Fe 2+ 34 , or ferric iron oxides (Fe 3 O 4 ) 35 . The relative lower BE peak at 712.5 eV is attributed to Fe 3+ , with a corresponding Fe 3+ satellite at 719.1 eV, which furthermore confirmed that both Fe 2+ and Fe 3+ were present in the nanocomposites. Figure 3(c) showed the O 1s XPS spectrum of A-ATP@Fe 3 O 4 . The spectrum can be fitted to four peaks with BEs of 530.6, 531.8, 532.6, and 533.5 eV. The peak at 530.6 eV resulted from the lattice oxygen in Fe 3 O 4 36 . The two peaks at 531.8 and 532.6 eV were attributed to the monodentate oxygen atoms (H-O) and monodentate and bidentate oxygen species (Si-O-Si), respectively 36,37 . The intensity ratios of the monodentate to bidentate oxygen  www.nature.com/scientificreports/ atoms showed that surface of A-ATP@Fe 3 O 4 was primarily bidentate. The remaining peak at 533.5 eV could be assigned to the chemically equivalent oxygen in the bidentate bond (O-C=O) 38 . For A-ATP@Fe 3 O 4 catalyst, chemisorbed oxygen is the most active oxygen species which plays an important role in the oxidation reaction. The Fourier transform infrared (FT-IR) spectra of the P-ATP, A-ATP, P-ATP@Fe 3 O 4 and A-ATP@Fe 3 O 4 samples are shown in Supplementary Fig. S2. In the spectrum of P-ATP and A-ATP, the absorbance bands at 3357 cm −1 and 3617 cm −1 were ascribed to the O-H stretching vibration of structural water and other water molecules in ATP 39 . The characteristic bands of stretching vibration of Si-O-Si for P-ATP and A-ATP was observed around 1033 cm −1 , as well as the bending vibration of H-O-H located at 1652 cm −1 . These three typical adsorption bands were also observed for P-ATP@Fe 3 O 4 and A-ATP@Fe 3 O 4 , but the absorption peaks were all weaker than those of the samples not loaded with Fe 3 O 4 , which implies that crystallization was essentially completely 40 . In addition, the peak at 582 cm −1 was owing to Fe-O bond for Fe 3 Table 1). The EtBr removal was ascribed mainly to the surface adsorption by ATP and Fe 3 O 4 minerals. The enhanced EtBr sorption by A-ATP was due to the increased surface area. The degradation reaction results (Fig. 4) showed that H 2 O 2 yielded only negligible removal of EtBr within 60 min. In the presence of H 2 O 2 , the degradation rate of EtBr using the Fe 3 O 4 -ATP composite was notably higher than that for Fe 3 O 4 , implying that the catalytic activity was enhanced by the introduction of ATP. After a 60 min heterogeneous Fenton reaction, Fe 3 O 4 -ATP composites exhibited a removal rate of 94%. In addition, the catalytic activity of A-ATP@Fe 3 O 4 composite was higher than those of P-ATP@Fe 3   ascribed to increased surface area of Fe 3 O 4 nanoparticles which were well dispersed on the surface of A-ATP. The relative rates of mass transfer to reactive sites and chemical reaction at reactive sites would thus be enhanced.
The pH effects on EtBr degradation by A-ATP@Fe 3 O 4 catalyst was determined ( Fig. 5 (a)). About 78% of EtBr was removed after 180 min of reaction at pH 5. EtBr degradation increased as pH decreased, suggesting that the production of •OH on the surface of A-ATP@Fe 3 O 4 was limited at higher pHs. Although EtBr removal rates decreased between pH 3 and 9, the nanocomposites still exhibited good EtBr degradation capacity. This implied that A-ATP@Fe 3 O 4 exhibited strong catalytic activity in a wide range of pH values. Under acidic, neutral, and alkaline conditions, the EtBr degradation was content with a pseudo first order reaction in kinetics, which might be expressed as ln(C t /C 0 ) = kt + m, where m is a constant, k is the apparent rate constant (min −1 ), C 0 is the residual concentration of EtBr (mmol/L) after 30 min absorption and C t is EtBr concentration at different sampling times. The highest k values of 2.445 min −1 for EtBr degradation was observed at pH 2.0, and thus pH 2.0 was selected for subsequent experiments.
The kinetics of EtBr degradation was also investigated at different temperatures (293, 303, 313, and 323 K). The activation energy (E a ) of the reaction was evaluated by plotting lnk against 1/T ( Fig. 5(b)) according to the Arrhenius equation. The activation energy was determined to be 78.39 kJ/mol for A-ATP@Fe 3 O 4 . This E a value falls within a reasonable range from the literature of 60 to 250 kJ/mol 44 . Dependence on the temperature in a heterogeneous Fenton-like reaction was previously reported through a carbon-Fe structured catalyst for the degradation of orange II with an activation energy of 56.1 kJ/mol (in a similar temperature range) 45 46 According to the classical Haber-Weiss mechanism 47 , Fe 2+ induces hydrogen peroxide to generate hydroxyl radicals (•OH), and the •OH can then react with Fe 3+ to regenerate Fe 2+ that can circularly produce •OH radicals in the Fenton reaction. However, k value declined to 0.446 min −1 at a higher H 2 O 2 concentrations of 40 mmol/L. It is possibly related to the scavenging effect of •OH radicals when excessive H 2 O 2 inhibits the production of •OH radicals 48 . To shorten reaction time, a higher concentration of H 2 O 2 (30 mmol/L) was applied for EtBr removal. In this case, it is necessary to investigate the loading of A-ATP@Fe 3 O 4 (Fig. 5(f)). In our study, as the amount of A-ATP@Fe 3 O 4 increased from 0.2 to 2.0 g/L, the rate constant k of EtBr degradation first increased and then decreased sharply. The increased removal rate may be due to the production of more reactive oxidants resulting from more active sites at higher rates of A-ATP@Fe 3 O 4 . The severe depression of EtBr removal is possibly ascribed to the scavenging of •OH radicals by excess Fe 2+ 49 . To conclude, better removal of EtBr with a shorter reaction time can be achieved under the following conditions: the 1.5 g/L of A-ATP@Fe 3 O 4 , pH = 2.0, T = 323 K, and 30 mmol/L H 2 O 2 (standard reaction condition).

Effect of Fe ion release.
To investigate the effects of the concentrations of dissolved Fe on the degradation of EtBr, the heterogeneous Fenton reaction was performed under the standard reaction condition. As shown in Fig. 6(a), in the adsorption stage, the concentration of ferrous ion increased gradually and reached a peak value of 2.48 mg/L, at which about 60% of EtBr was absorbed. After H 2 O 2 was added, the concentration of Fe 2+ decreased to about 0.26 mg/L where the removal rate of EtBr was 90% after 20 min. The reason is that the catalyst can release ferrous ions to the acid solution, and H 2 O 2 can oxidize the ferrous ions to generate •OH 50 . Different from other heterogeneous Fenton reaction, a small amount of ferrous ion was released from the catalyst in the catalytic degradation stage. However, the fast degradation rate at 2 min implies that H 2 O 2 exhibited an excellent ability to oxidize ferrous ions to produce •OH quickly, leading to a fast decreasing of ferrous ions. Further, the variation of ferrous ions also caused the increase of the dissolved iron from the A-ATP@Fe 3 O 4 composite and the oxidation of ferrous ions in solution. The dissolved iron amounted to 5 mg/L, which is equivalent to about 0.62% of total iron in the catalyst (1.5 g/L).

Reusability of A-ATP@Fe 3 O 4 . The spent A-ATP@Fe 3 O 4 was recycled and reused for EtBr degradation
under standard reaction condition. As shown in Fig. 6(b), A-ATP@Fe 3 O 4 maintained more than 80% of its catalytic capacity after five successive runs in 150 min of reaction. The reduced EtBr degradation efficiency probably resulted from the reduction of released iron ions from the catalyst in each successive runs. Therefore, to maintain an adequate quantity of the catalyst in the aqueous solutions and thus maintain the degradation efficiency, a prolonged degradation time may be needed.

Possible degradation mechanisms. The EtBr degradation process was indicated by mineralization
(reduction of TOC) (Fig. 7(a)) and the UV-visible absorption spectrum. The results indicated that the maximum TOC removal rate was approximately 45% after 20 min, suggesting that 45% of EtBr was oxidized by active species(•OH, •O 2 − , etc.) to CO 2 and H 2 O. This result is confirmed by the UV-vis absorption spectra of EtBr ( Fig. 7(b)), which showed that the characteristic peak at λ = 285 nm became smaller and almost disappeared as the degradation proceeded. In addition, we observed that the absorption peak at λ = 241 nm weakened sharply after the addition of H 2 O 2 , indicating that the attack by the highly reactive hydroxyl radicals led to rapid opening of the benzene ring 51 . In addition, a new absorption peak at λ = 210 nm was recorded after 2 min of treatment and decreased gradually with reaction time. This may suggest that EtBr was attacked by hydroxyl radicals to produce a large number of intermediate products 52 , although further confirmation is needed.
To confirm reactive species in the degradation process, t-butyl alcohol and benzoquinone (BQ) were selected as radical scavengers during EtBr degradation, respectively. Excess t-butyl alcohol could scavenger all of the •OH produced by the system in the solution 53 . As shown in the Fig. 7(c), the degradation reaction rate of EtBr decreased evidently after the addition of t-butyl alcohol, which indicated the existence of •OH. However, about 20% of EtBr was not affected by the presence of t-butyl alcohol, which suggests the existence of other reactive species. After that, BQ was added to the system as a scavenger of •O 2 − 54 . From Fig. 7(c), with the addition of excess BQ, the EtBr degradation decreased from 68%, 77%, 86%, and 90% (in the absence of BQ) to 56%, 69%,    Figure 7(d,II) clearly shows that the six characteristic peaks of DMPO-•O 2 − existed in the degraded system. Combining the results from the reactive oxygen species assay and the EPR analysis, we determined that •OH and •O 2 − were present in the catalytic system. Thus, the possible reaction mechanism of H 2 O 2 activation by A-ATP@Fe 3 O 4 under acidic condition is illustrated in Fig. 8. According to our observations, it is possible that Fe 2+ and Fe 3+ from partial dissolution of iron oxides under acidic conditions initiate the decomposition of H 2 O 2 through a homogeneous Fenton chain reaction. Initially, the dissolved Fe 2+ can react with H 2 O 2 to generate Fe 3+ and •OH (equation (1)), which yield HO 2 •/•O 2 − and simultaneously produce Fe 2+ (equation (2)). The generated •OH (equation (1)) may further react with H 2 O 2 to generate HO 2 •/•O 2 − (equation (3)). Then partial Fe 2+ is converted to Fe 3+ through oxidation of the produced •OH (equation (4)). Thus, all of these reactions lead to cycling of iron ions in the solution system. In our work, a large number of H + ions in solution facilitate that the reactions in equations (1) and (4), leading to decreased Fe 2+ along with increased Fe 3+ during the catalytic reactions.
Moreover, the amount of EtBr degradation declined from 68%, 77%, 86%, and 90% to 51%, 59%, 64%, and 66% at 2, 5, 10, and 15 min in the presence of excess KI (10 mM), suggesting that •OH originating from the surface of the catalyst played a dominant role in EtBr degradation (Fig. 7(c)). And the total amount of iron dissolved remained constant ( Fig. 6(a)), implying that not much Fe 2+ was released in EtBr degradation. Therefore, we conjecture that similar chain reactions occurred on the surface of the catalyst, which is different from the that in solution. Moreover, the hydroxyl radicals generated on the catalyst surface by the chain reaction are also involved in the degradation of EtBr. This may explain the reduction of catalytic performance in the process of reusing, which is similar with the result reported by Luo et al. 55 .  Preparation and characterization of catalyst. The pristine ATP powder was purified as follows. Briefly, 20 g of ATP were dispersed in 300 mL of deionized water and stirred for 2 h. The resulting slurry was then settled for 2 h, and the supernatant was decanted to remove impurities. Next, the prepared ATP was immersed in 300 mL of deionized water with 200 mL of H 2 O 2 (30%, w/w). The suspension was magnetically stirred for 5 h, and sonicated for 30 min (40 kHz). The suspension was centrifuged at 5000 rpm, and the resulting precipitate was vacuum dried for 24 h at 70 °C and stored for subsequent use. P-ATP was then activated in 150 mL of 1 M HCl with constant magnetic stirring for 5 h followed by sonication for 30 min. A-ATP was vacuum filtered, washed with deionized water and ethanol to a pH value of 6-7, and then dried for 24 h at 70 °C under vacuum.The prepared ATP was stored in a desiccator at room temperature for further use.
A-ATP@Fe 3 O 4 were synthesized by co-precipitation method 56 . A-ATP (2.0 g) were added to 240 mL of deionized water in a 500 mL flask and stirred for 6 h. Then, the pH of the suspension was adjusted to 8 using 5 M NaOH solution. The stable suspension was bubbled with a constant N 2 flow for 30 min to remove the dissolved oxygen 57 . Next, a 0.6 mol/L FeSO 4 solution (20 m L) was added to the flask and sonicated for 30 min, followed by addition of 0.8 mol/L FeCl 3 solution (20 ml). The mixture was then sonicated for 15 min. NaOH solution (5 M) was added to the flask drop wisely. Black precipitate appeared immediately after NaOH addition, and the reaction was terminated at a pH of 10. The nanocomposites were aged at 60 °C for another 1.5 h. The suspension was centrifuged at 4000 rpm for 5 min; the as-prepared Fe 3 O 4 loaded ATP was washed with deionized water and ethanol several times to remove free ions and dried in a vacuum oven at 70 °C for 24 h. Finally, A-ATP@Fe 3 O 4 nanoparticles with an Fe 3 O 4 -to-ATP mass ratio of 1:1 were obtained. In addition, P-ATP@Fe 3 O 4 was prepared following the above procedure without HCl activation and Fe 3 O 4 alone was synthesized without adding ATP. All the products were stored in a desiccator under room temperature before use.
Characterization of catalyst. The morphology of the catalyst was observed on a scanning electron microscope (S-4800II, Japan) operated at an acceleration voltage of 15 kV. The surface groups of the nanocomposites were recorded by a micro infrared spectrometer (Cary 610/670, USA). The phase structure of the nanocomposites was obtained by XRD analysis (D8 Advance Bruker AXS, Germany). EDS were measured using a X-ray energy dispersive spectrometer (Thermo Electron Corporation) with Al Kα radiation as the excitation source. XPS (ESCALAB 250 Xi, USA) was used to identify the metal oxidation states of the nanocomposites. The magnetization of ATP-Fe 3 O 4 and Fe 3 O 4 was measured at room temperature using vibrating sample magnetometry (VSM-EV7, ADE) with a maximum applied field of 1.7 T. The specific surface area of the catalysts was determined by N 2 -BET analysis using an accelerated surface area and a porosimetry analyzer (ASAP 2460).

Degradation experiment.
Batch degradation experiments of EtBr were conducted in a conical flask (250 mL) incubated in a water bath with a constant temperature oscillator (TZ-2EH, Beijing Wode Co.) and shaken at 150 rpm in darkness. The reaction suspension was prepared by adding the required amount of catalyst (0.2-2.0 g/L) to 200 mL of an 80 mg/L EtBr solution at different pH values (2.0-9.0). The suspension was vibrated for 30 min to achieve the adsorption/desorption equilibrium. The EtBr concentration after equilibrium was measured and considered as the initial concentration (C 0 ). Then, a known concentration of H 2 O 2 was added to initiate the degradation reaction. Subsamples were taken at set intervals during the reaction using a 3 mL centrifuge tube and immediately centrifuged at 5000 rpm for 5 min using an H1650-W centrifuge (HuNan) to remove the catalyst. The EtBr concentration of the supernatant was determined at λ = 285 nm by using a UV-visible spectrophotometer. Each experiment was run in triplicate.
To test the regeneration ability of Fe 3 O 4 -ATP, spent nanoparticles were separated from the suspension when the EtBr was almost completely degraded. The regenerated sorbents were then used again for EtBr degradation. The regeneration process was repeated five times.
The presence of hydroxyl radicals and superoxide radical were determined using t-butyl alcohol, KI and BQ, respectively, as scavengers. The effective radicals that appeared in the degradation process were further detected by electron spin resonance (A300-10/12, Bruker, Germany).The concentration of ferrous ions was measured colorimetrically with 1,10-phenanthroline at λ = 510 nm on a UV-vis spectrophotometer 58 . The total dissolved iron was analyzed by atomic absorption spectroscopy (G8433A).