Preparation of Fe 3 O 4 / TiO 2 / C Nanocomposites and Their Application in Fenton-Like Catalysis for Dye Decoloration

Fe-H2O2 Fenton system is widely applied in water treatment nowadays, but the acidification and sludge generation are crucial problems to be solved. Herein, we report that Fe3O4/TiO2/C nanocomposites (FTCNCs) were able to catalyze the decomposition of H2O2 at neutral pH and can be applied in dye decoloration. FTCNCs were prepared by precipitating TiO2 on Fe3O4 cores via the hydrolysis of tetrabutyl titanate followed by the hydrothermal dehydrogenization of glucose to deposit carbon on Fe3O4/TiO2. The decoloration of methylene blue (MB) in the FTCNC-H2O2 Fenton-like system was monitored to reflect the catalytic activity of FTCNC. The radical generation capability was analyzed by electron spin resonance. Our results indicated that FTCNC-H2O2 Fenton-like system was efficient in decolorizing MB, and the radicals led to the near complete oxidation of MB. The FTCNC-H2O2 Fenton-like system could be used in a wide pH range of 4–9. A greater catalyst amount, a higher H2O2 concentration, and a higher temperature accelerated the decoloration kinetics. FTCNCs showed good activity after the regeneration of 8 cycles. The implication to the practical applications of FTCNCs in water treatment is discussed.


Introduction
Advanced oxidation processes (AOPs) are important technologies for water treatment, in particular for the pollutants that are hardly biotransformed in active sludge treatment.Generally, oxidative radicals or other intermediates are generated during AOP treatment and the oxidative species oxidize the electron-enriched organic pollutants [1][2][3].Among the most commonly applied AOPs, the Fe 2+ -H 2 O 2 Fenton system has attracted great interest and has been used for water remediation on a large scale.Despite the success in practical applications, the Fe 2+ -H 2 O 2 Fenton system has several marked disadvantages that require further improvements.The most vital one is that the optimal pH of the Fe 2+ -H 2 O 2 Fenton system is acidic (at pH 3).Thus, polluted water has to be acidified before treatment and followed by neutralization after treatment.Another drawback is that the Fe 2+ -H 2 O 2 Fenton system requires a stoichiometric amount of iron.The iron ions are hard to remove and generate a high degree of sludge [4,5].The iron containing sludge should be further treated to avoid environmental hazards.
In recent years, heterogeneous Fenton-like catalysis has been developed as a new alternative to the traditional Fe 2+ -H 2 O 2 Fenton system.Typically, heterogeneous Fenton-like catalysts are microsized or nanosized and contain Fe or other metal elements, where the surface metal atoms catalyze the decomposition of H 2 O 2 .Many heterogeneous catalysts have been reported to catalyze the decomposition of H 2 O 2 for water treatment [6][7][8][9][10][11][12].Both iron-containing and iron-free catalysts have been studied, and external energy supply (e.g., sonication [13,14], electricity [2] and light irradiation [9,[15][16][17]) is adopted in many cases to enhance the performance of heterogeneous catalysis.For instance, humic acid modified Fe 3 O 4 served as a Fenton-like catalyst and catalyzed the degradation of sulfathiazole [6].Fe 3 O 4 /CeO 2 nanocomposites catalyzed the degradation of methyl orange under microwave irradiation [8].The MnO 4 microstructure could be used for the degradation of methylene blue (MB) [18].To simplify the conditions, much research has been dedicated to developing nanoparticles as Fenton-like catalysts without an external energy supply.However, such attempts usually worked under acidic pH levels (around pH 3).
Previously, we reported the doping of TiO 2 on a Fe 3 O 4 nanoparticle surface to shift the applicable pH to near-neutral [19].However, the stability and regeneration of TiO 2 -doped Fe 3 O 4 nanoparticles were poor.Here, we prepared Fe 3 O 4 /TiO 2 /C nanocomposites (FTCNC) as a Fenton-like catalyst for the decomposition of H 2 O 2 and tested the performance of the FTCNC-H 2 O 2 system in the decoloration of MB.FTCNCs were carefully characterized before used, and the catalytic performance was compared with naked Fe 3 O 4 at a neutral pH level.The influences of radical scavengers, catalyst amount, H 2 O 2 concentration, pH, and temperature were investigated.The regeneration capability was analyzed up to 8 cycles.The implication on the applications of FTCNCs in water treatment is discussed.

Characterization of FTCNCs
FTCNCs were characterized by multiple techniques before use.The morphology of FTCNCs was investigated under transmission electron microscopy (TEM, Figure 1).It was recognized that the particles were irregular.The size distribution of particles was very wide with large particles of about 20 nm in diameter and small ones of about 5 nm.The particles aggregated to form larger composites due to the magnetic attraction and the high surface energy, even though the aggregate was loosely packed and the pores were presented under TEM, which would allow the diffusion of small molecules onto the inner surface of FTCNC.The surface area was measured by the Brunauer-Emmett-Teller (BET) technique, and the value was 96.5 m 2 /g.The large surface area was consistent with the small particle sizes.The N 2 adsorption/desorption isotherm curve of FTCNCs followed type IV (Figure 2a), indicating the adsorption hysteresis nature.The magnetic property measurement showed the ferromagnetic character of FTCNC, which should be due to the nanoscale sizes.The saturated magnetization of FTCNCs was 48.5 emu/g, similar to other modified Fe 3 O 4 nanoparticles.The ferromagnetic character was good for the applications of FTCNC, because FTCNCs could disperse better without an external magnet during the catalysis and could be magnetically separated after the reaction with the external magnet.
The chemical composition was analyzed by X-ray photoelectron spectroscopy (XPS).O (43.5 atom %), C (38.6 atom %), Fe (14.0 atom %), and Ti (3.9 atom %) were detected in FTCNC.The low Ti content indicated that the TiO 2 only partially covered the surface of Fe 3 O 4 , consistent with the TEM observation, where no core-shell structure was presented.The contents of both Fe and Ti reduced after the hydrothermal reaction, suggesting that Fe and Ti were partially lost during the reaction to form some nonmagnetic substances.The XPS spectra of C, Fe, and Ti are shown in Figure 3.According to the analyses, the carbon atoms could be divided into C-C (56%), C-O (34%), and C=O (10%).The C1s spectrum suggested that the glucose was carbonized, but many oxygen atoms were retained.The existence of Ti 4+ was confirmed by the typical Ti 2p 3/2 and Ti 2p 1/2 peaks.The Fe 3+ /Fe 2+ ratio was calculated from the peak areas of Fe 2p 3/2 as 1.35:1, smaller than those of Fe 3 O 4 (2:1) and the starting Fe 3+ /Fe 2+ ratio (1.5:1), likely due to the chemical reduction during the hydrothermal reaction.The groups containing oxygen, distinguished from the C1s XPS spectrum, was confirmed by the infrared (IR) spectrum.The peak at 3390 cm −1 is attributed to -OH and -COOH groups.The tiny peak at 2970 cm −1 was due to the remnant C-H bonds on the residues of glucose.The graphite carbon (C=C) is reflected by the peak at 1620 cm −1 .The peaks at 579 cm −1 and 432 cm −1 were assigned to Ti-O and Fe-O.hydrothermal reaction.The groups containing oxygen, distinguished from the C1s XPS spectrum, was confirmed by the infrared (IR) spectrum.The peak at 3390 cm −1 is attributed to -OH and -COOH groups.The tiny peak at 2970 cm −1 was due to the remnant C-H bonds on the residues of glucose.The graphite carbon (C=C) is reflected by the peak at 1620 cm −1 .The peaks at 579 cm −1 and 432 cm −1 were assigned to Ti-O and Fe-O.hydrothermal reaction.The groups containing oxygen, distinguished from the C1s XPS spectrum, was confirmed by the infrared (IR) spectrum.The peak at 3390 cm −1 is attributed to -OH and -COOH groups.The tiny peak at 2970 cm −1 was due to the remnant C-H bonds on the residues of glucose.The graphite carbon (C=C) is reflected by the peak at 1620 cm −1 .The peaks at 579 cm −1 and 432 cm −1 were assigned to Ti-O and Fe-O.

Decoloration of MB in the FTCNC-H2O2 Fenton-Like System
The FTCNC-H2O2 Fenton-like system showed good capability in decolorizing MB (Figure 4a), suggesting the high catalytic activity of FTCNCs in the decomposition of H2O2.The decoloration was carried out at near-neutral pH (pH 6) [19].In contrast, very limited decoloration was observed

Decoloration of MB in the FTCNC-H2O2 Fenton-Like System
The FTCNC-H2O2 Fenton-like system showed good capability in decolorizing MB (Figure 4a), suggesting the high catalytic activity of FTCNCs in the decomposition of H2O2.The decoloration was carried out at near-neutral pH (pH 6) [19].In contrast, very limited decoloration was observed

Decoloration of MB in the FTCNC-H 2 O 2 Fenton-Like System
The FTCNC-H 2 O 2 Fenton-like system showed good capability in decolorizing MB (Figure 4a), suggesting the high catalytic activity of FTCNCs in the decomposition of H 2 O 2 .The decoloration was carried out at near-neutral pH (pH 6) [19].In contrast, very limited decoloration was observed in the Fe 3 O 4 -H 2 O 2 system and the catalyst-free system.The decoloration efficiency of MB was 82% for the FTCNC-H 2 O 2 Fenton-like system, 9.4% for the Fe 3 O 4 -H 2 O 2 system and 6.7% for the catalyst-free system.The results clearly indicate that the co-incorporation of TiO 2 and C shifted the working pH of Fe 3 O 4 to near-neutral.We also analyzed the kinetics of decoloration.The corresponding kinetics constant k values were calculated as 0.011 min −1 for the FTCNC-H 2 O 2 Fenton-like system, 0.00055 min −1 for the Fe 3 O 4 -H 2 O 2 system, and 0.00031 min −1 for the catalyst-free system.The k value for the FTCNC-H 2 O 2 Fenton-like system was two magnitudes higher than the other two.The kinetics of the FTCNC-H 2 O 2 -MB system was competitive with other systems using various nanocatalysts.For example, the Fe 3 O 4 @SiO 2 -H 2 O 2 -MB system had a k value of 0.020 min −1 [7].The k values of the FePt-H 2 O 2 -MB system were in the range of 0.0033-0.023min −1 at the pH value of 5.5 [20].The Ti-doped Fe 3 O 4 -H 2 O 2 -MB system had a similar k value (0.0165 min −1 ).From the perspective of kinetics, the incorporation of carbon did not improve the performance of TiO 2 -doped Fe 3 O 4 [19].The major improvements induced by carbon incorporation were the sample stability and regeneration.The TiO 2 -doped Fe 3 O 4 was vulnerable during storage, while FTCNCs could be stored under room temperature for more than two months.The regeneration of FTCNCs was much better than the TiO 2 -doped Fe 3 O 4 .Moreover, we also measured the chemical oxygen demand (COD) to verify the complete oxidation of MB (Figure 4c).The trends of the optical density and COD decreases were nearly identical, which suggested that the decoloration of MB was mainly due to the complete oxidation of MB into CO 2 [20][21][22][23][24][25]. in the Fe3O4-H2O2 system and the catalyst-free system.The decoloration efficiency of MB was 82% for the FTCNC-H2O2 Fenton-like system, 9.4% for the Fe3O4-H2O2 system and 6.7% for the catalyst-free system.The results clearly indicate that the co-incorporation of TiO2 and C shifted the working pH of Fe3O4 to near-neutral.We also analyzed the kinetics of decoloration.The corresponding kinetics constant k values were calculated as 0.011 min −1 for the FTCNC-H2O2 Fenton-like system, 0.00055 min −1 for the Fe3O4-H2O2 system, and 0.00031 min −1 for the catalyst-free system.The k value for the FTCNC-H2O2 Fenton-like system was two magnitudes higher than the other two.The kinetics of the FTCNC-H2O2-MB system was competitive with other systems using various nanocatalysts.For example, the Fe3O4@SiO2-H2O2-MB system had a k value of 0.020 min −1 [7].The k values of the FePt-H2O2-MB system were in the range of 0.0033-0.023min −1 at the pH value of 5.5 [20].The Ti-doped Fe3O4-H2O2-MB system had a similar k value (0.0165 min −1 ).From the perspective of kinetics, the incorporation of carbon did not improve the performance of TiO2-doped Fe3O4 [19].The major improvements induced by carbon incorporation were the sample stability and regeneration.The TiO2-doped Fe3O4 was vulnerable during storage, while FTCNCs could be stored under room temperature for more than two months.The regeneration of FTCNCs was much better than the TiO2-doped Fe3O4.Moreover, we also measured the chemical oxygen demand (COD) to verify the complete oxidation of MB (Figure 4c).The trends of the optical density and COD decreases were nearly identical, which suggested that the decoloration of MB was mainly due to the complete oxidation of MB into CO2 [20][21][22][23][24][25].Generally, many Fenton systems generate oxidative radicals and other non-radical species [26].To confirm the existence of •OH radicals in the FTCNC-H 2 O 2 Fenton-like system, the electron spin resonance (ESR) spectra of the FTCNC-H 2 O 2 Fenton-like system were recorded [21].As shown in Figure 5, the 4-fold characteristic peaks of 5,5-dimethyl-1-pyrroline-N-oxide (DMPO)-•OH adduct was observed with the intensity ratio of 1:2:2:1 in the ESR spectra of the FTCNC-H 2 O 2 Fenton-like system.More radicals were generated at 10 min than those at 5 min, which was reflected by the higher intensity.The radicals could be partially quenched by tertiary butanol, a typical scavenger for radicals (Figure 5b).Along with the increase of tertiary butanol concentration, the catalytic activity of FTCNCs was inhibited.However, it should be noted that, even at very high tertiary butanol (0.909 mol/L, 3878 times of that of MB), the decoloration was still meaningful with a k value of 0.0066 min −1 .The resistance of the FTCNC-H 2 O 2 system to tertiary butanol should be due to the incorporation of TiO 2 and C. It was speculated that radicals were generated on the surface of the FTCNCs, and the radicals then diffused from the surface to the solution phase, forming higher radical concentrations around the surface of the FTCNCs.MB interacted more intensively with the FTCNCs than tertiary butanol; thus, more MB was concentrated around the surface of the FTCNCs and reached the radicals prior to tertiary butanol [21].Therefore, radicals oxidized MB before they diffused into the solution phase, where the extinction of radicals by tertiary butanol occurred.The tolerance of the FTCNC-H 2 O 2 system to radical scavengers was important for practical applications, because there might be radical scavengers in polluted water, and no pre-purification is required.Beyond the radical process, according to the literature, there might also be non-radical processes involved.Recently, van Eldik et al. found a new mechanistic aspect of Fenton reactions in which not only hydroxyl radicals but also Fe III ( − O 2 H) aq were involved [26].Thus, H 2 O 2 decomposed to generate several oxidative species that further oxidized the MB molecules.
Generally, many Fenton systems generate oxidative radicals and other non-radical species [26].To confirm the existence of •OH radicals in the FTCNC-H2O2 Fenton-like system, the electron spin resonance (ESR) spectra of the FTCNC-H2O2 Fenton-like system were recorded [21].As shown in Figure 5, the 4-fold characteristic peaks of 5,5-dimethyl-1-pyrroline-N-oxide (DMPO)-•OH adduct was observed with the intensity ratio of 1:2:2:1 in the ESR spectra of the FTCNC-H2O2 Fenton-like system.More radicals were generated at 10 min than those at 5 min, which was reflected by the higher intensity.The radicals could be partially quenched by tertiary butanol, a typical scavenger for radicals (Figure 5b).Along with the increase of tertiary butanol concentration, the catalytic activity of FTCNCs was inhibited.However, it should be noted that, even at very high tertiary butanol (0.909 mol/L, 3878 times of that of MB), the decoloration was still meaningful with a k value of 0.0066 min −1 .The resistance of the FTCNC-H2O2 system to tertiary butanol should be due to the incorporation of TiO2 and C. It was speculated that radicals were generated on the surface of the FTCNCs, and the radicals then diffused from the surface to the solution phase, forming higher radical concentrations around the surface of the FTCNCs.MB interacted more intensively with the FTCNCs than tertiary butanol; thus, more MB was concentrated around the surface of the FTCNCs and reached the radicals prior to tertiary butanol [21].Therefore, radicals oxidized MB before they diffused into the solution phase, where the extinction of radicals by tertiary butanol occurred.The tolerance of the FTCNC-H2O2 system to radical scavengers was important for practical applications, because there might be radical scavengers in polluted water, and no pre-purification is required.Beyond the radical process, according to the literature, there might also be non-radical processes involved.Recently, van Eldik et al. found a new mechanistic aspect of Fenton reactions in which not only hydroxyl radicals but also Fe III ( − O2H)aq were involved [26].Thus, H2O2 decomposed to generate several oxidative species that further oxidized the MB molecules.Beyond the good resistance to radical scavengers, the FTCNC-H2O2 Fenton-like system was capable of working at different pH levels and temperatures.We investigated the influence of four factors, namely catalyst amount, H2O2 volume, pH, and temperature.As described in Figure 6, more FTCNCs and H2O2 accelerated the decoloration.This was reasonable from the kinetics perspective and widely observed in heterogeneous Fenton reactions [8].At a high catalyst amount, the increase of decoloration kinetics slowed down, which suggested that the diffusion of MB toward the active sites of FTCNCs controlled the reaction kinetics.The optimal pH for the FTCNC-H2O2 Fenton-like system was pH 6.The FTCNC-H2O2 Fenton-like system also worked in the pH range of 4-9.A slightly acidic environment (pH 4-6) seemed much better than neutral and basic environments (pH 7-9).The wide applicable pH range implied that the FTCNC-H2O2 Fenton-like system can be used at various pH values without pH adjustment.This obviously overcame the main drawback of the Fe 2+ -H2O2 system, which only worked at pH 3. The shift of optimal pH toward neutral should be attributed to the doping of TiO2 on Fe3O4.Previously, we have demonstrated that TiO2 doping could push the applicable pH of the Fenton reaction to near-neutral [19], while C coating did not change the optimal pH level.According to the experiences of chelators and SiO2 [21], possible mechanism could be that -OH groups of TiO2 chelated with Fe 2+ and the chelation shifted the optimal pH level.It is well known that the point of zero charge (PZC) of TiO2 was near pH 6 [27].At a pH level higher than 6, the hydroxyl groups on TiO2 deprotonated to form -O − , making the surface negative.By reducing the pH to 6, hydroxyl groups were protonated to form -OH and the surface charge was neutral.Further reduction of pH generated -OH2 + , making the surface positive.At a low pH level, the positive charge of -OH2 + repelled MB toward the surface due to electrostatic repulsion.At a high pH level, the lack of proton inhibited the Fenton-like reaction.At the optimal pH, close to the PZC of TiO2, the status of surface hydroxyl groups was -OH, which can chelate with Fe 2+ .By increasing the temperature, the decoloration was significantly accelerated.The increase in temperature accelerated the diffusion of MB toward active sites and more molecules reached the activation energy [28].By fitting the kinetics constant and temperature to the Arrhenius equation (R = 0.995), the activation energy of decoloration was 26.35 kJ/mol.The small activation energy was consistent with the fast decoloration.Beyond the good resistance to radical scavengers, the FTCNC-H 2 O 2 Fenton-like system was capable of working at different pH levels and temperatures.We investigated the influence of four factors, namely catalyst amount, H 2 O 2 volume, pH, and temperature.As described in Figure 6, more FTCNCs and H 2 O 2 accelerated the decoloration.This was reasonable from the kinetics perspective and widely observed in heterogeneous Fenton reactions [8].At a high catalyst amount, the increase of decoloration kinetics slowed down, which suggested that the diffusion of MB toward the active sites of FTCNCs controlled the reaction kinetics.The optimal pH for the FTCNC-H 2 O 2 Fenton-like system was pH 6.The FTCNC-H 2 O 2 Fenton-like system also worked in the pH range of 4-9.A slightly acidic environment (pH 4-6) seemed much better than neutral and basic environments (pH 7-9).The wide applicable pH range implied that the FTCNC-H 2 O 2 Fenton-like system can be used at various pH values without pH adjustment.This obviously overcame the main drawback of the Fe 2+ -H 2 O 2 system, which only worked at pH 3. The shift of optimal pH toward neutral should be attributed to the doping of TiO 2 on Fe 3 O 4 .Previously, we have demonstrated that TiO 2 doping could push the applicable pH of the Fenton reaction to near-neutral [19], while C coating did not change the optimal pH level.According to the experiences of chelators and SiO 2 [21], possible mechanism could be that -OH groups of TiO 2 chelated with Fe 2+ and the chelation shifted the optimal pH level.It is well known that the point of zero charge (PZC) of TiO 2 was near pH 6 [27].At a pH level higher than 6, the hydroxyl groups on TiO 2 deprotonated to form -O − , making the surface negative.By reducing the pH to 6, hydroxyl groups were protonated to form -OH and the surface charge was neutral.Further reduction of pH generated -OH 2 + , making the surface positive.At a low pH level, the positive charge of -OH 2 + repelled MB toward the surface due to electrostatic repulsion.At a high pH level, the lack of proton inhibited the Fenton-like reaction.At the optimal pH, close to the PZC of TiO 2 , the status of surface hydroxyl groups was -OH, which can chelate with Fe 2+ .By increasing the temperature, the decoloration was significantly accelerated.The increase in temperature accelerated the diffusion of MB toward active sites and more molecules reached the activation energy [28].By fitting the kinetics constant and temperature to the Arrhenius equation (R = 0.995), the activation energy of decoloration was 26.35 kJ/mol.The small activation energy was consistent with the fast decoloration.

The Recycle of FTCNCs
The regeneration of catalyst after decoloration is essential in reducing the operating cost and environmental burden.As shown in Figure 7, the regeneration led to a slight decrease in decoloration efficiency of the FTCNC-H2O2 Fenton-like system.After the first round of recycle, the decoloration efficiency kept nearly constant, indicating that some vulnerable active sites of FTCNCs were lost in the first round.The remnant sites were much more stable and survived in the following seven rounds of recycle.Compared with the TiO2-doped Fe3O4 [19], FTCNCs showed much better regeneration capability.The improvement of regeneration should be attributed to the incorporation of carbon.Previously, the sacrificial role of graphene oxide (GO) was confirmed in the regeneration of GO-Fe3O4 [29].The carbon atoms of GO were oxidized during the Fenton reaction; therefore, Fe 2+ remained non-oxidized.A similar phenomenon has been observed in Fe3O4/SiO2/C nanocomposites, where carbon incorporation enhanced the regeneration of Fe3O4/SiO2 [20].Overall, upon the incorporation of carbon, FTCNCs had good regeneration ability and is thus more promising for practical applications [21,28].

The Recycle of FTCNCs
The regeneration of catalyst after decoloration is essential in reducing the operating cost and environmental burden.As shown in Figure 7, the regeneration led to a slight decrease in decoloration efficiency of the FTCNC-H 2 O 2 Fenton-like system.After the first round of recycle, the decoloration efficiency kept nearly constant, indicating that some vulnerable active sites of FTCNCs were lost in the first round.The remnant sites were much more stable and survived in the following seven rounds of recycle.Compared with the TiO 2 -doped Fe 3 O 4 [19], FTCNCs showed much better regeneration capability.The improvement of regeneration should be attributed to the incorporation of carbon.Previously, the sacrificial role of graphene oxide (GO) was confirmed in the regeneration of GO-Fe 3 O 4 [29].The carbon atoms of GO were oxidized during the Fenton reaction; therefore, Fe 2+ remained non-oxidized.A similar phenomenon has been observed in Fe 3 O 4 /SiO 2 /C nanocomposites, where carbon incorporation enhanced the regeneration of Fe 3 O 4 /SiO 2 [20].Overall, upon the incorporation of carbon, FTCNCs had good regeneration ability and is thus more promising for practical applications [21,28].

Preparation of FTCNCs
FeCl2•4H2O (0.45 g) and FeCl3•6H2O (0.90 g) were dissolved in 40 mL of deionized water.The pH was adjusted to pH 12 with a NaOH solution under vigorous stirring.After aging for 1 h, the magnetic components were collected and washed with deionized water.The as-prepared Fe3O4 was dried for TiO2 doping.A portion of 0.1 g Fe3O4 was sonicated in a mixture of 20 mL of alcohol and 100 μL of NH3•H2O for 5 min.A mixture of 1.62 mL of tetrabutyl titanate and 20 mL of alcohol was added dropwise into the Fe3O4 suspension.The suspension was then shaken at 100 rpm at 35 °C for 6 h.The TiO2-doped Fe3O4 was magnetically separated, washed, and dried.Then, 0.4 g of TiO2-doped Fe3O4, 1.6 g of glucose, and 60 mL of deionized water were placed in a Teflon tube (Hengli Co., Chengdu, China) and heated to 160 °C for 6 h.The precipitates were magnetically separated to collect FTCNCs.FTCNCs were washed and lyophilized before characterization and catalysis.FTCNCs were characterized by TEM (Tecnai G2 20, FEI, Hillsboro, OR, USA), the BET technique (ASAP2010, Micromeritics, Norcross, GA, USA), a magnetometer (MPMS XL-7 tesla, Quantum Design, San Diego, CA, USA), XPS (Kratos, Stretford, UK), and an IR spectrometer (Magna-IR 750, Nicolet, Rhinelander, WI, USA).

Decoloration of MB in FTCNC-H2O2 Fenton-Like System
20 mg of FTCNCs were added to a 20-mL MB solution (75 mg/L, pH 6) and shaken at 110 rpm at 35 °C for 2 h to reach the adsorption equilibrium.The absorbance of the MB solution was recorded at 664 nm on a UV-vis spectrometer (UV-1800, Mapada, Shanghai, China) as A0.Then, 1 mL of H2O2 was added to the system (the final concentration of H2O2 was 0.48 mol/L) and was kept shaking.At each interval of 20 min, the absorbance was recorded as At.The decoloration efficiency was calculated with Equation (1).The decoloration kinetics were analyzed with Equation (2).The corresponding COD values of water samples were measured using a Hach reagent (low range 3-150 mg/L) on the Hach DR900 (Hach Co., Loveland, CO, USA).For comparison, the decoloration of MB in the Fe3O4-H2O2 system and the catalyst-free system was determined following the same protocol.

Preparation of FTCNCs
FeCl 2 •4H 2 O (0.45 g) and FeCl 3 •6H 2 O (0.90 g) were dissolved in 40 mL of deionized water.The pH was adjusted to pH 12 with a NaOH solution under vigorous stirring.After aging for 1 h, the magnetic components were collected and washed with deionized water.The as-prepared Fe

Decoloration of MB in FTCNC-H 2 O 2 Fenton-Like System
20 mg of FTCNCs were added to a 20-mL MB solution (75 mg/L, pH 6) and shaken at 110 rpm at 35 • C for 2 h to reach the adsorption equilibrium.The absorbance of the MB solution was recorded at 664 nm on a UV-vis spectrometer (UV-1800, Mapada, Shanghai, China) as A 0 .Then, 1 mL of H 2 O 2 was added to the system (the final concentration of H 2 O 2 was 0.48 mol/L) and was kept shaking.At each interval of 20 min, the absorbance was recorded as A t .The decoloration efficiency was calculated with Equation (1).The decoloration kinetics were analyzed with Equation (2).The corresponding COD values of water samples were measured using a Hach reagent (low range 3-150 mg/L) on the Hach DR900 (Hach Co., Loveland, CO, USA).For comparison, the decoloration of MB in the Fe 3 O 4 -H 2 O 2 system and the catalyst-free system was determined following the same protocol.

Figure 5 .
Figure 5. Resistance of the FTCNC-H 2 O 2 Fenton-like system against radical scavenger tertiary butanol.(a) ESR spectra of FTCNC-H 2 O 2 Fenton-like system at 5 and 10 min; (b) influence of tertiary butanol.
3 O 4 was dried for TiO 2 doping.A portion of 0.1 g Fe 3 O 4 was sonicated in a mixture of 20 mL of alcohol and 100 µL of NH 3 •H 2 O for 5 min.A mixture of 1.62 mL of tetrabutyl titanate and 20 mL of alcohol was added dropwise into the Fe 3 O 4 suspension.The suspension was then shaken at 100 rpm at 35 • C for 6 h.The TiO 2 -doped Fe 3 O 4 was magnetically separated, washed, and dried.Then, 0.4 g of TiO 2 -doped Fe 3 O 4 , 1.6 g of glucose, and 60 mL of deionized water were placed in a Teflon tube (Hengli Co., Chengdu, China) and heated to 160 • C for 6 h.The precipitates were magnetically separated to collect FTCNCs.FTCNCs were washed and lyophilized before characterization and catalysis.FTCNCs were characterized by TEM (Tecnai G2 20, FEI, Hillsboro, OR, USA), the BET technique (ASAP2010, Micromeritics, Norcross, GA, USA), a magnetometer (MPMS XL-7 tesla, Quantum Design, San Diego, CA, USA), XPS (Kratos, Stretford, UK), and an IR spectrometer (Magna-IR 750, Nicolet, Rhinelander, WI, USA).