A facile preparation of g-C3N4/a-Fe2O3/TiO2 composite and its photocatalysis performance on the degradation of DNT/TNT under simulated sunlight

It has been difficult to solve the environmental problems caused by trinitrotoluene(TNT)/dinitrotoluene(DNT) containing explosive wastewater for decades. The advanced oxidation method based on photocatalysis and hydrogen peroxide has gained considerable popularity in the field of pollutant degradation. In order to find a coping strategy, using a simple method involving obtaining g-C3N4/Fe2O3(CNFe) by two-step calcination and then mixing butyl titanate and CNFe through a sol-gel pathway, a type of low-cost g-C3N4/Fe2O3/TiO2(CNFeTi) ternary composite was prepared and then characterized by SEM-EDS, IR, Raman spectroscopy, XRD, XPS, UV-vis, etc. Using the prepared CNFeTi as catalyst, the effects of the amount of catalyst and H2O2 dosage on the photocatalytic degradation of DNT/TNT in aqueous solution have been investigated under simulated sunlight, with the reaction being tracked by UV-vis spectrophotometer. And from the results it was shown to have excellent photocatalytic performance. Under the conditions of 1.0 g·L−1 H2O2, 3.0 g·L−1 C3N4/Fe2O3/TiO2, 15 A current intensity of a Xenon light source and natural pH, 95% DNT and 91% TNT were removed after 30 min irradiation, indicating the proposed system has a high degradation rate. The CNFeTi catalyst could be reused at least five times and exhibited good stability.


Introduction
The large-scale production and application of nitroaromatic compounds (NACs), such as mononitrotoluene, 2,4-dinitrotoluene, 2,6-dinnitrotoluene (DNT), and 2,4,6-trinitrotoluene (TNT), by human industrial and military activities have caused contamination of the water and soil environment with high potential ecological and human health risks [1,2] .And it is difficult to remove NACs by natural degradation.Therefore, to avoid environmental hazards, NACs must be removed before discharge.The prevention and management of the NACs pollution have been the focus of worldwide attention for decades.A common drawback of using traditional wastewater treatment techniques, including physical, chemical and biotechnologies for the treatment of such wastewater and their resulting contamination, is the removal of low efficiency or high treatment costs and possible secondary contaminations [3,4] .Nowadays, the heterogeneous photocatalytic technology has the practical application potential [5][6][7] with the economic, efficient and environmental-friendly photo-catalysts and the developed photocatalytic processes, and will play an important role in this field.
Many valuable works have been reported in the photocatalytic treatment of wastewater containing mono-, di-or trinitrotoluene using TiO2 and its composites as catalysts [8][9][10][11][12][13][14][15][16][17][18][19][20][21][22][23][24][25] , which covered the aspects of the construction of catalyst, photocatalytic reaction conditions, wavelength range and intensity of the light source, reaction mechanism and dynamics, influence of H2O2, using combined technologies, and so on.Schmelling et al [11,12] treated TNT wastewater with TiO2 suspension and fixed membranes, the TNT degradation rate was up to 95% after 1 h of light irradiation, and the recovery of photocatalysts was successfully resolved.Son et al [17] observed that TNT was completely degraded in TiO2/UV system at 90 min with a higher degradation rate than that of either UV light or TiO2 alone.Shashi B. Mishra et al [18] studied the adsorption of TNT on the rutile TiO2(110) surface and found that the TNT adsorption is dependent on the vacancy concentration [18] .Using liquid-liquid extraction and mass spectrometry, the presence of intermediates like 2,4-dinnitrobenzyl formaldehyde, 2,6dinnitrobenzyl formaldehyde and dinitro-benzoic acid was demonstrated during the photo-degradation of DNT [22,23] .Similar reaction path and product distribution were observed by Patience C. Ho [8] .All of the above partly resolve the photocatalytic degradation pathway of DNT/TNT.And the photocatalytic degradation process of DNT was consistent with quasi-primary dynamics when supplemented with deionized water and UV light.In conclusion, the photocatalytic degradation study of TNT and DNT mainly adopts the TNT/DNT degradation of TiO2 or TiO2-based composite photo-catalyst under UV irradiation.The application of photo-catalyst is limited, and the degradation mechanism needs to be expanded, therefore, this aspect still needs to be studied, including the comprehensive use of sunlight photo-catalyst, efficient photocatalytic reaction system and the corresponding photocatalytic degradation reaction mechanism.Photo-catalysis is a complex chemical process wherein the photo-catalyst absorbs photon energy to generate electron(e -) and hole(h + ) pairs which migrate to the catalyst surface to initiate redox reaction of molecules adsorbed to the catalyst surface.During efficient photo-degradation, quantum efficiency must be improved, extending the excitation wavelength range of photo-catalysts to make full use of sunlight.Resovskite type TiO2 is the most studied photo-catalyst driven by UV light [24,26] , but its broad band gap (3.2eV) can only be excited by 388 nm UV light.Therefore, other visible photocatalysts or oxide composite TiO2 is required to obtain higher photocatalytic performance under visible light applications.It was well known that, g-C3N4 with a band gap (2.7ev), acts photocatalytically against water decomposition and organic matter degradation under visible light exposure [27-29]   , which is easy to be prepared, a soft polymer facilitating the formation of a hybrid structure that play an important role on the separation efficiency of photogenic (e -) and (h + ) pairs [30][31][32] .Fe2O3, as a rich, eco-friendly metal oxide, with a band gap (2.3 eV) and visible-light photocatalytic activity [33] , was reported to exhibit good photocatalytic performance of Fe2O3 composites or dopants with g-C3N4 [34][35][36] .The g-C3N4/Fe2O3 coupling complex improves the electron-hole pair separation efficiency during light irradiation, while increasing the application of visible light.It was also found that an appropriate amount of Fe2O3 incorporated with TiO2 showed better UV photocatalytic performance than TiO2 alone, with the self-built electric field formed in the hetero-structures of Fe2O3 and TiO2 benefitting to UV absorption, providing more electrons for the catalytic process and improving the catalytic properties [37] .Based on the above, the combination of g-C3N4, TiO2 and Fe2O3 with different band gaps will improve the separation efficiency of (e -) and (h + ) pairs by heterogeneous section, and synergistically enhance the sunlight-driven photocatalytic activity.Therefore, after preliminary experimental exploration, we designed and prepared a type of a-Fe2O3/g-C3N4/TiO2 (CNFeTi) composite, a low-cost ternary complex, and used it to study the photocatalytic degradation of DNT/TNT under simulated sunlight irradiation(xenon lamp).Based on the method [34] for constructing g-C3N4/Fe2O3 (CNFe) with heterojunctions by co-calcination of melamine and Fe(NO3)3, the ternary CNFeTi composite was synthesized by simple sol-gel method with Ti(OBu)4 and CNFe.Subsequently, the catalysts were characterized by FTIR, SEM-EDX, PXRD, XPS, and solid UV-Vis diffuse reflection spectroscopy(UV-Vis DRS), the effects of the catalyst dosage, reaction time and H2O2 have been investigated on the photocatalytic degradation of DNT/TNT, and the reaction mechanism was inferred.

Preparation of g-C3N4, Fe2O3
, and g-C3N4/Fe2O3.g-C3N4: Yellow g-C3N4 powder was prepared by heating treatment of melamine by copying the method of reference [30] .g-C3N4/Fe2O3 (CNFe): Solution of 2 g Fe(NO3)3•9H2O in 50 mL of anhydrous ethanol, added with melamine (5 g) under continuous stirring, and dissolved at 40 ℃ for 2 h.It was then distilled under reduced pressure at 60 ℃ for 1 h to remove ethanol.The resulting solid was put into a vacuum drying oven at 60 ℃ for 2.5 h.Subsequently, the solid was heated to 500 ℃ in a crucible with a lid in N2 atmosphere at 5 ℃/min and holding for 1 h, then continued to 520 ℃ at 5 ℃/min and holding for 1 h.Then the oven was naturally cooled to room temperature.And the products were obtained and stored for use after grinding.Fe2O3 was prepared from Fe(NO3)3•9H2O at 500 ℃ for 3 h in a furnace as above, finely ground, sealed and stored for comparison experiments.g-C3N4/Fe2O3 /TiO2(CNFeTi): At room temperature, anhydrous ethanol(20 mL), Ti(OBu)4(3 mL) and deionized water(3 mL) were added to a 50 mL beaker, and stirred at 40 ℃ to produce white precipitation, then 0.1 mL of nitrate acid (67%) dropped into the turbid solution and mixed for 30 min, and sol A was obtained.At the same time, 0.4 g CNFe powder and 50 mL of anhydrous ethanol were added into a three-necked bottle, stirred for 15 min and then heated to 80 ℃ (with cooling reflux), then sol A was added dropwise, the system was kept refluxing for 30 min.Then the ethanol was removed by reduced pressure distillation at 60 ℃, the solid products were cleaned with anhydrous alcohol and deionized water for 3~5 times and then put into a vacuum drying oven for 10 h.Then the CNFeTi complex was obtained.

Photocatalytic degradation of DNT/TNT.
It was carried out in a photo-reactor with a PLS-SXE300 / UV300 Xenon light source with a current set to 15 A, which was temperature-controlled by flowing water.For a typical process, DNT (100 mg•L -1 ) or TNT(50 mg•L -1 ) aqueous solution was prepared with deionized water at 40 ℃ for 48 h.50 mL of DNT solution and an amount of catalyst were added to a 50 mL beaker, an amount of H2O2 was added as designed.After magnetically stirred at 25 ℃ for 30 min in the dark, the system was sampled 1 mL for the first time, using a 5 mL syringe with micro-filter membrane (0.2 μm).Then this beaker was transferred to the photo-reactor to start the photocatalytic reaction under the xenon light irradiation.The light path length was about 10 cm to the liquid level.The sample was then diluted in 2 m L of 5% ethanol aqueous solution and centrifuged for 5 min at 3000 rpm and then the UV-vis analysis followed.The reaction system was sampled every 5 or10 min as designed until the photocatalytic reaction was terminated.The experiments on TNT were carried out in a similar manner.The effects of the amount of the ternary catalyst CNFeTi and the dosage of hydrogen peroxide on DNT photocatalytic degradation were investigated, and recycled use of CNFeTi was subsequently performed to comprehensively assess the activity of the photo-catalyst.

Characterization of the catalysts
The functional group of the material was determined on a FTIR spectrometer (Tensor 27 type Fourier conversion infrared spectrometer), where the KBr powder was used as a background material and detected by conventional pressing method to obtain infrared spectroscopy.The structure and morphology of the catalysts were observed on a field emission scanning electron microscope (FESEM) under 10.00 KV.Surface elemental contents of the catalysts were analysed by performing an energy dispersion analysis using an X-ray (EDX) on the same SEM instrument.Photo activity analysis was carried out with UV-Vis DRS to obtain light absorbent wavelength range and energy band gap of the prepared catalyst.XPS test was carried out to obtain the elements and electronic orbitals and valence states of the catalyst.

Analysis of the photocatalytic degradation of DNT/TNT waste water
Using a UV-1800 Isadzu UV spectrophotometer, the absorbance A of DNT/TNT waste water at 252 nm was measured to tracking the change of the DNT/TNT concentration in the photocatalytic system.That is, the percentage of current DNT concentration to the initial is calculated by At /A0, where At is the absorbance of the sample of DNT/TNT waste water at a specific time, and A0 represents the absorbance at an initial concentration.With At/A0 as the ordinate and the time as the abscissa, the curve of photocatalytic degradation efficiency was plotted.

Figure 1. Infrared spectrogram of catalysts
As shown in Figure 1, the curve of pure g-C3N4 showed vibration of aromatic C-N bonds at bands of 1425,1327 and 1250 cm -1 , C=N vibration at 1637 cm -1 , and the tri-s-triazine ring at 808 cm -1 [32] .From the curve of Fe2O3, the peaks at 538 and 478 cm -1 represent the scaling vibration of Fe-O [34] .Obviously, the curves of CNFe and CNFeTi showed the same characteristic adsorption bands of g-C3N4 as in the curve of pure g-C3N4, indicating the same structure of g-C3N4 in the three materials.The absorption at band 500-800 cm -1 represented the tensile vibration of Fe-O and Ti-O, indicating relative low contents of Fe2O3 or TiO2 in the CNFe and CNFeTi composites, respectively.And as the doping component increases, the IR absorption intensity became weaker, and some peaks migrate slightly towards the high wavenumber zone due to the interaction between the components at their interfaces [30]   .

SEM and EDS
SEM and EDS results of g-C3N4, CNFe, and CNFeTi were shown in Figure 2. Figure 2 (a), (b), (c) and (d) showed that g-C3N4 has a planar lamellar structure, and CNFe presented features like shells with holes and smaller particle sizes, which will contribute to TiO2 loading.The EDS results showed that the atomic content of C in g-C3N4 surface was about 31.21%, and the N content approximately 68.79%.The average atomic content of C in CNFe surface was about 30.97%,N about 44.51%, O about 17.26%, and Fe about 7.27%.The results indicated that heterojunctions between g-C3N4 and Fe2O3 have been formed in the CNFe complex.The distribution of TiO2 particles on the surface of CNFe particles could be observed from Figure 2 (e), (f) and (g).Interestingly, the TiO2 particles supported by CNFe exhibited better dispersion than the nearby stacked TiO2 particles.It also showed the close contact of TiO2 particles with the surface of CNFe, and CNFe may play a role to support the exposure of TiO2 particles, thus providing more active catalytic reaction sites.The results of the EDS analysis showed that the average element atomic contents in the surface of CNFeTi were about 20.05% of C, 42.34% of N, 29.22% of O, 1.79% of Fe, and 6.61% of Ti.The special morphology and structure of the CNFeTi composite should facilitate the effective separation of photogenic (e -) -(h + ) pairs, thus improving the photocatalytic performance.

PXRD
Figure 3 showed the PXRD patterns of as-prepared catalysts.As shown in the figure, the composite CNFeTi has the characteristic diffraction peaks of g-C3N4, Fe2O3, CNFe, and TiO2, although the intensity of the peaks is somewhat weak, which may be attributed to the formation of new crystalline phases and crystalline surfaces in the composite, which indicating the formation of hetero-phase interfaces.

XPS
The structure and chemical states of the composite photo-catalyst CNFeTi were determined by XPS.The total spectrum and high resolution spectra of the XPS of CNFeTi were shown in Figure 4.
As could be seen from the total spectrum of the composite, all the elements of C, N, O, Fe and Ti were found, one main peak of C 1s was at the binding energy of about 285 eV, one N 1s peak was about 399 eV, and one O 1s peak was about 530 eV, and one peak of Ti 2p was about 458 eV, and one peak of Fe 2P was about 710 eV, proving that the three units in the CNFeTi composite were well combined.
In the high resolution spectra of Ti, two peaks of Ti 2p3/2 and 2p1/2 were detected at about 458 eV and 464 eV, respectively, It is proved that the composite photocatalyst contains the Ti element, and that the Ti element exists in the + 4 valence state.The binding energy of Ti (2p1/2) is reduced in compare with that Ti (2p1/2) is 464.8 eV in pure TiO2, which might be attributed to the interaction between the complex components, which increases the electron cloud density of Ti, resulting in the weakening of atomic bonds and binding energy, which would benefit the electron transfer and thus improving the catalytic performance [38] .

UV-vis DRS
The results were showed in Figure 5.

Figure5. UV-vis DRS analysis and the estimated band gaps of the catalysts.
As can be seen from the Figure, pure g-C3N4 displayed the absorption edge at about 440 nm.The prepared Fe2O3 and CNFe showed a wide and strong light absorption from UV to visible light region, between them, Fe2O3 presented stronger.CNFeTi presented stronger light absorption in UV region and weaker in visible area than CNFe.The estimated Eg of the CNFe and CNFeTi were 2.31 eV and 2.41 eV, respectively.The results indicated that the ternary composite CNFeTi could response to the UVvis light irradiation when under simulated sunlight.

Effect of catalyst species on DNT degradation
The photocatalytic performance of each catalyst was studied with 100 mg•L -1 of the DNT waste water.The amount of various catalysts in the study was uniform to 6 g•L -1 .Five samples were collected at 3 min intervals and every 12 min as one round.Figure 6 (a) showed the degradation rate of DNT by various catalysts.It could be observed that the DNT is degraded by about 6% within 12 min with absence of any catalyst in the photo-degradation system, presented a slow degradation rate.The six catalysts of g-C3N4, Fe2O3, Fe2O3 (500 ℃), TiO2, CNFe, and CNFeTi showed about 21%, 35%, 31%, 15%, 23%, 43% degradation percent, respectively.Figure 6 (b) shows the degradation efficiency of various catalysts in a system containing a hydrogen peroxide concentration of 0.2 g•L -1 .DNT was degraded by approximately 15% within 12 minutes under the light experimental conditions, which shows that hydrogen peroxide promotes the photo-degradation of DNT.The six catalysts of g-C3N4, Fe2O3, Fe2O3 (500 ℃), TiO2, CNFe, and CNFeTi showed degradation of about 20%, 23%, 27%, 18%, 21%, 48%.The ternary catalyst designed and prepared shows excellent photo-degradation performance, and with the addition of hydrogen peroxide, it most likely acts as an oxidant to generate •OH involved in the degradation reaction, and accelerating the destruction of DNT.The photocatalytic activity of iron oxide is secondary.As a kind of raw material, low-cost pigment, used as a photocatalyst also shows a certain efficiency.Whether purchased directly or calcinated with furnace, the degradation rate has reached about 30%.However, the addition of H2O2 did not aggravate the decomposition of DNT, and it is speculated that there is no synergy between iron oxide and hydrogen peroxide, and it cannot activate H2O2 to effectively degrade DNT.For TiO2, g-C3N4, CNFe, the first two were outstanding representatives of photo-catalysts, and considerable reports have been reported on themselves and derived studies; the latter is an important work active in the field in recent years, namely the study of carbon and nitrogen-doped metal oxides to overcome the shortcomings of existing photocatalysts to achieve higher catalytic activity.However, experiments have shown that they performed generally when used to degrade DNT, and could not change such results after the addition of hydrogen peroxide.It could be seen that they also had no obvious synergy with hydrogen peroxide under these conditions.

Effect of g-C3N4/Fe2O3/TiO2 catalyst dosage on DNT degradation
In similar methods as above, the ternary catalyst CNFeTi used in a quantity of 0.05 g, 0.10 g, 0.15 g, 0.20 g exhibited DNT degradation by about 22%, 23%, and 37%, 24% degradation rate in 15 min, respectively.It could be seen that the more amount of catalyst did not necessarily represent a better catalytic performance.However, the dosage of 0.25 g, 0.30 g exhibited about 43% and 48% degradation percent, respectively.Their reaction rate could be seen from Figure 7 to follows a certain constant under different dosage, except that 0.15 g is relatively special.For the phenomenon of different reaction rate, it was speculated that the comprehensive influence was based on the number of surface active sites of the photocatalyst, light-borne charge carrier transport, the light transmittance, the effective utilization rate of the photocatalyst on the light, and the distribution of the photocatalyst in the degradation system, and the interactivity between the catalysts surface and the substrate.
Although the amount of 0.30 g CNFeTi exhibited the highest degradation rate in this group, considering the absolute weight of DNT (5 mg) in a reaction round, in fact, each increase of 0.05 g for the final degradation efficiency rate increased from 0.15g to 0.30g.Therefore, the high degradation rate shown at 0.30 g consumption may not be enough to compensate for the increased costs due to the increased consumption.With a relative dosage of 0.15 g, it also has an acceptable degradation performance, while the cost was easily controlled, and the degradation rate is relatively medium and conducive to sampling monitoring, which should be the next choice dosage to study other influencing factors.

Effect of H2O2 dosage on the degradation of DNT/ TNT
The effect of H2O2 dosage on the photo-degradation of DNT/TNT was investigated on the basis of the aforementioned studies.The amount of CNFeTi was 0.15 g.When H2O2 dosage was 0.2, 0.6, 0.8, 1.0, 1.4 g•L -1 , it showed a degradation of about 50%, 76%, 85%, 95.3%, and 96.2% in 30 min, respectively (Figure 8).It could be seen that as H2O2 dosage increased, the degradation percent of DNT increased, reaching nearly 100% at 1.0 g•L -1 H2O2.When the H2O2 dosage was up to 1.4 g•L -1 , the increase in the degradation rate of DNT became very small.The photo-degradation rate of catalyst-free solution of TNT solution (about 50 mg•L -1 ) was studied and showed that TNT degraded about 8% within 30 min under experimental light, which indicated the difficult degradability of TNT.The degradation rate of 50 mL TNT solution (concentration of about 50 mg•L -1 ) was used to study that of 0.15 g of CNFeTi and 1.0 g•L -1 H2O2).The results showed that TNT was degraded by 91% within 30 min of light irradiation.This suggested that the CNFeTi photocatalyst could be used to effectively remove nitroaromatic compounds, including DNT and TNT.

Kinetics of the degradation process
The kinetic parameters of the degradation process were calculated based on a quasi-primary reaction model.The relationship between ln (Cr/Co) and time was plotted in Figure 9, showing that the DNT/TNT degradation process follows a linear relationship of quasi-primary reaction kinetics except for TNT catalyst-free degradation.Obviously, in Figure 9, there showed a highly nonlinear relationship between ln (Cr/Co) and time during the whole degradation stage of TNT degradation and after DNT degradation for 20 minutes.The classical first-level kinetic model could not better describe the degradation process, and the actual process was complicated and needed to be further studied.

Recycling of the CNFeTi
The stability and reusability of CNFeTi were observed by recycling experiments.As shown in Figure 10, DNT waste water still showed degradation rate above 75% after five cycles of the CNFeTi photocatalyst.Furthermore, no significant changes were found between the fresh catalyst sample and five-cycled one based on FTIR spectrum, indicating good stability of the CNFeTi catalyst.

The speculated photocatalytic mechanism
A possible mechanism of CNFeTi with high photocatalytic efficiency was proposed by consideration all above.Figure 11 showed the separation of (e -) -(h + ) pairs, transfer processes, and synergistic effects within the three units under simulated sunlight irradiation.And the process still need to be studied by catalyst surface species capturing, quantum chemical calculations and other means.
Sometimes, surface defects and heterojunctions might have a crucial role on the catalytic properties, as shown in this study.The excellent catalytic performance of the CNFeTi might be attributed to the synergy and the tight heterointerface among the g-C3N4, Fe2O3 and TiO2 [39].

Conclusion
In conclusion, the g-C3N4/Fe2O3/TiO2(CNFeTi) composite was constructed on considering the synergy among the three units.The morphological structure and optical properties of the CNFeTi were characterized and the effects of CNFeTi on the DNT/TNT photocatalytic degradation were investigated.Using an amount of catalyst with mixing appropriate dosage of hydrogen peroxide under simulated sunlight irradiation, CNFeTi showed the most effective photocatalytic activity.The degradation rate of DNT reached more than 95%, while TNT reached about 91%.The CNFeTi could be recycled at least five times with good stability.Comparing pure TiO2 and g-C3N4, the improved photocatalytic activity was probably attributed to effective charge separation, suitable band location, abundant adsorption and active sites on the catalyst surface, the synergy among the g-C3N4, Fe2O3 and TiO2.

Figure 6 .
Figure 6.(a) degradation efficiency of DNT by different types of catalysts and (b) degradation efficiency of different types of catalysts by DNT at a H2O2 concentration of 0.2 g•L -1 .

Figure 7 .
Figure 7. Photo-catalytic degrade efficiency of DNT waste water by CNFeTi

Figure 8 .
Figure 8.Effect of H2O2 dosage on the photo-degradation of DNT/ TNT.It could be seen that as H2O2 dosage increased, the degradation percent of DNT increased, reaching nearly 100% at 1.0 g•L -1 H2O2.When the H2O2 dosage was up to 1.4 g•L -1 , the increase in the degradation rate of DNT became very small.The photo-degradation rate of catalyst-free solution of TNT solution (about 50 mg•L -1 ) was studied and showed that TNT degraded about 8% within 30 min under experimental light, which indicated the difficult degradability of TNT.The degradation rate of 50 mL TNT solution (concentration of about 50 mg•L -1 ) was used to study that of 0.15 g of CNFeTi and 1.0 g•L -1 H2O2).The results showed that TNT was degraded by 91% within 30 min of light irradiation.This suggested that the CNFeTi photocatalyst could be used to effectively remove nitroaromatic compounds, including DNT and TNT.

Figure 11 .
Figure 11.A possible mechanism of DNT/TNT degradation by CNFeTi photo-catalysis under simulated sunlight