Photocatalytic degradation of toxic aquatic pollutants by novel magnetic 3D-TiO2@HPGA nanocomposite

In this study, a series of photocatalysts were prepared, namely bare 3D-TiO2 (b-3D-T), magnetic 3D-TiO2: (m3D-T) and magnetic 3D-TiO2@Hierarchical Porous Graphene Aerogels (HPGA) nanocomposite: (m3D-T-HPGA NC) by solvothermal process. The prepared photocatalysts were analyzed by using X-ray diffraction (XRD), Field emission scanning electron microscope (FE-SEM), transmission electron microscope (TEM), Vibrating sample magnetometer (VSM), Brunauer–Emmett–Teller (BET) and Diffuse Reflectance Measurement – Ultraviolet (DRS-UV) to know their physical and chemical properties. The photocatalytic degradations of two toxic aquatic pollutants viz., Cr(VI) and bisphenol A (BPA) were tested by using the prepared photocatalysts. Parameters such as initial pollutant concentration, solution pH, photocatalyst dosage, wavelength and light intensity were investigated to optimize the process. The photocatalytic properties of prepared catalyst were analyzed based on the degradation of Cr(VI) and BPA under UV irradiation. The modified photocatalysts showed better performance as compared to b-3D-T photocatalyst. This better performance is ascribed to efficient charge transfer between b-3D-T nanoparticles to the porous graphene sheets. The maximum photocatalytic degradation of Cr(VI) was found to be 100% with m3D-T-HPGA NC within 140 min, whereas a removal efficacy of 100% and 57% was noticed in case of m3D-T and b-3D-T within 200 and 240 min, respectively. In the case of BPA, the maximum degradation efficiency was found to be 90% with m3D-T-HPGA NC within 240 min.


Results and Discussion
Characterization of photocatalysts. XRD   ferrite (CoFe 2 O 4 ), which is in good agreement with the standard JCPDS No: 22-1086 23 . Fig. S1(a) shows the XRD analysis of bare CoFe 2 O 4 nanoparticles which is in good agreement with the above crystal planes and diffraction peaks. Figure 1(c) shows the XRD analysis of m3D-T-HPGA NC. All the diffraction peaks and crystal planes are matched with the m3D-T nanocomposite. No diffraction peak of hierarchal porous graphene aerogel (HPGA) was observed in the prepared nanocomposite. Fig. S1(b) shows (002) and (101) crystal plane at 2θ of 26.7° and 41.2° diffraction peaks respectively, which confirms the presence of graphene hydrogel peak. The presence of (002) and (101) was disappeared in the nanocomposite (Fig. 1(c)) because graphene sheets were decorated on the surface of the b-3D-T and CoFe 2 O 4 nanoparticles and the detection of thin graphene layers was difficult 24 . Hence, the prepared nanocomposites were synthesized without any impurities by using the solvothermal process. Figure 2(a) shows the FE-SEM images of b-3D-T material synthesized in this study. Flower like structures of TiO 2 were seen from the images (Fig. 2(a)). Each TiO 2 flower like structures consists of nanorods in hierarchical form possessing a microsphere. Figure 2(b) shows a zoom-in view of flower like 3D-TiO 2 microspheres. Each individual sphere possesses an average size of 1.5 µm, whereas each nanorod was found to have an average diameter of 20-30 nm with a length of 300-350 nm. As seen from the images, each individual nanorod was self-assembled to form a 3D-TiO 2 microsphere with a high compact hierarchical structure. With such a well confined structure, it could enhance the photocatalytic properties by facilitating the charge carriers transport easily. Figure 2(c) shows the SEM image of m3D-T nanocomposite, which reveals that as compared to b-3D-T, the magnetic particles (CoFe 2 O 4 ) were well anchored on the surface of 3D-TiO 2 . The average particle size of the magnetic nanoparticles was in the diameter of 20-25 nm. As clearly seen from the images (Fig. 2(d)), the tips of 3D-TiO 2 nanorods were covered fully by the CoFe 2 O 4 nanoparticles. Hence the prepared nanocomposite was successfully synthesized by the solvothermal process. Finally, Fig. 2(e,f) shows the m3D-T-HPGA NC. As compared to the m3D-T, CoFe 2 O 4 nanoparticles were well decorated without any agglomeration. This might be explained due to the incorporation of HPGA on to the m3D-T, and as a result, graphene layers help to increase the surface to volume ratio as well as avoiding the agglomeration between the magnetic nanoparticles. Figure 3 shows the TEM images with EDAX of b-3D-T and bare HPGA nanomaterials. Figure 3(a-d) clearly shows the flower like 3D-TiO 2 spheres with EDAX analysis. From the figures, it can be clearly seen that a bunch of nanorods (1D) were formed to a flower like structure (3D). The diameter of each nanorod was approximately in the range of 20-25 nm size. The nanorod looks like 1D structures (Fig. 3b), and these bunch of nanorods then form 3D flower like TiO 2 structures ( Fig. 3(a,c)). Figure 3d shows the EDAX analysis of b-3D-T, which reveals the elements present in the material. The major element present in the b-3D-T nanomaterial is titanium (Ti) (47.59 in wt.%) and oxygen (O) (12.48 wt.%). Some other elements are also present in the sample such as copper, which arises from the TEM copper grids. Figure 3(e-h) clearly shows the graphene flakes with few number of sheets with a d-spacing of 0.347 nm. Figure 3(h) shows the EDAX of HPGA and the major elements present in the material are carbon (C) and oxygen (O). The carbon is in the highest amount (99.01 wt.%), whereas oxygen is 0.22 wt.% and the remaining is copper which arises from TEM copper grids. Figure S2 shows the TEM images with EDAX of m3D-T and m3D-T-HPGA NC. Figure S2(a-d) clearly shows that the flower like 3D-TiO 2 spheres were fully decorated by the magnetic (CoFe 2 O 4 ) nanoparticles. From the figures, it can be clearly seen that all the 3D flower like TiO 2 spheres were well decorated with the magnetic nanoparticles. Fig. S2(d) shows the EDAX analysis of m3D-T, which reveals the elements present in the material. The major elements present in the m3D-T nanomaterial are oxygen (O) (35.89 in wt.%), titanium (Ti) (24.4 in wt.%), iron (Fe) (32.08 in wt.%) and cobalt (Co) (7.64 in wt.%). Some other elements are also present in the sample such as copper, which arises from the TEM copper grids. Fig. S2(e-h) shows the m3D-T-HPGA NC with EDAX analysis, and it can be seen from the figures that the m3D-T nanomaterial was decorated on graphene flakes.  Table 1.  Figure 4a(i) shows the M-H hysteresis of b-3D-T spheres, which is almost zero because TiO 2 is a non-magnetic material. Inset figure also shows that the saturation magnetization value of TiO 2 is 0.02 emu/g, which is negligible and equals to zero. Figure 4a(ii) shows the M-H hysteresis of m3D-T nanocomposite. This material shows 33 emu/g of saturation magnetization, which resembles that the material is well coated with the CoFe 2 O 4 nanoparticles onto the 3D-TiO 2 spheres. Magnetic 3D-TiO 2 @HPGA (m3D-T-HPGA NC) also shows the good magnetic property with a saturation magnetization of 32 emu/g ( Fig. 4a(iii)), (note that HPGA is also a non-magnetic material). Hence, there is slight difference between the m3D-T and m3D-T-HPGA NC. Overall, the hysteresis loop analysis suggests that the material is well decorated with the magnetic nanoparticles (CoFe 2 O 4 ) onto the 3D-TiO 2 and HPGA nanomaterials with a super paramagnetic in nature. This is very important for recovering and reuse of the material from water and wastewater after the process.

VSM analysis.
Optical properties. The diffuse reflectance UV-visible spectrum of b-3D-T, m3D-T and m3D-T-HPGA NC are shown in Fig. 4(b). The prepared b-3D-T exhibits the higher absorbance in UV region and the maximum absorption peak arises at about 354 nm, whereas in the case of m3D-T and m3D-T-HPGA NC, there was a broad peak in the region of 300 to 800 nm. The broad peak was observed due to the blue shift of the absorption edges of m3D-T and m3D-T-HPGA NC as compared to b-3D-T. Using Kubelka-Munk function 25 , the band gap energy values were calculated for the prepared nanocomposites, which are shown in Fig. 4(c). The plot shows the intercept line towards the x-axis, which gives the bandgap energy value of the prepared nanocomposites. The b-3D-T photocatalyst shows the bandgap energy of 3.05 eV and whereas m3D-T and m3D-T-HPGA NCs show the bandgap energy of 1.5 eV. The bandgap energy is low for the two nanocomposites as compared to b-3D-T material, which may be due to the introduction of magnetic nanoparticles and HPGA materials to the TiO 2 material. The change in size and shape might have caused a change in the band gap of the material. The addition of magnetic particles will lead to easy recovery of the nanomaterial from the water, whereas the role of HPGA is to increase the surface area and it will have more recombination rates, which will finally result in the degradation of the pollutants within the specified time. Figure 5 shows the Mott-Schottky plots for the m3D-T and m3D-T-HPGA NC. As shown in Fig    surface area, which is due to hierarchical porous graphene aerogel (HPGA) with meso-pore structures. By using Barrett-Joyner-Halenda (BJH) model, the pore distribution of the prepared nanocomposites was analyzed and shown in Fig. S3b. The prepared m3D-T-HPGA showed that the average pore size distribution is 14.25 nm.
Photoelectrochemical studies. The photocurrent and electrochemical impedance spectroscopy (EIS) are vital characterization techniques to investigate the separation and migration of charge carriers in the photocatalysts. Figure 6 shows the photocurrent response of b-3D-T, m3D-T and m3D-T-HPGA NC with three on-off cycles under UV -irradiation. From the photocurrent results ( Fig. 6(a)), it can be observed that m3D-T-HPGA NC exhibited a higher photocurrent intensity as compared to m3D-T and b-3D-T, which reveals that the charge separation efficiency in m3D-T-HPGA NC was higher than other catalysts and hence, it can generate more photo   induced charge carriers. In addition, EIS measurement was performed to investigate the charge transfer resistance of the material. Figure 6(b) shows the Nyquist plot of b-3D-T, m3D-T and m3D-T-HPGA NC, respectively. As seen from Fig. 6(b), m3D-T-HPGA NC possesses a smaller arc radius than the m3D-T and b-3D-T. The smaller arc radius of m3D-T-HPGA NC implies the lower at the contact interface. This further proves that the interfacial charge-transfer and separation efficiency of photo excited charge carriers is higher for m3D-T-HPGA NC which is much valuable for its superior photocatalytic activity.
Photocatalytic degradation studies. Effect of photocatalyst dosage. Figure 7(a,b) shows the effect of photocatalyst dosage on the degradation of Cr(VI) with respect to the degradation ratio (C/C o ) and various intervals of time (t). Figure 7(a) shows the photodegradation of 10 mg L −1 Cr(VI) ions with a catalyst dosage of 0.1 g L −1 under pH 2 with a wavelength of 254 nm and with 4 lamps intensity (32 W). Cr(VI) concentration was found to reduce to some extent in the absence of photocatalyst. This suggests that UV light has some minor effect on the degradation of Cr(VI). However, in the presence of b-3D-T, degradation of Cr(VI) was observed (50-60% in 240 min). Furthermore, the photocatalytic activity of m3D-T shows a significant degradation of Cr(VI) ions, and within 200 min 100% degradation of Cr(VI) was achieved. Moreover, in the case of m3D-T-HPGA NC, a 100% degradation of Cr(VI) ions was achieved within 140 min. The mechanism of photocatalytic activity is schematically represented in Fig. S4 and briefly explained as follows. When UV light is illuminated on the photocatalyst (3D-TiO 2 ), electron hole pair is generated, where the electrons are excited at its conduction band and holes in its valency band. The reduction of Cr(VI) to Cr(III) subsequently occurs due to photoexcited electrons (e − ), and water molecule is oxidized to O 2 , due to the holes (h + ) generated in the valency band. The degradation efficiency was found to be two times higher for m3D-T-HPGA NC than b-3D-T, which can be explained due to the fact that m3D-T-HPGA possesses high specific surface area and enhanced generation of electron-hole pairs. Figure 7(b) shows the photodegradation of 10 mg L −1 Cr(VI) ions with a catalyst dosage of 0.2 g L −1 under similar conditions. Among the three photocatalysts, m3D-T-HPGA NC shows the best results and 100% degradation of Cr(VI) ions was achieved within 100 min under UV irradiation. Fig. S5(a) shows that the prepared nanocomposites followed the pseudo first order kinetic model for Cr(VI) ions 26 .   Figure 7(d) shows the photodegradation of 10 mg L −1 BPA with a catalyst dosage of 0.2 g L −1 under similar conditions. Among the three photocatalysts, m3D-T-HPGA NC showed the best results and almost 90% degradation of BPA was achieved within 240 min under UV irradiation. Fig. S5(b) shows that the prepared nanocomposites followed the pseudo first order kinetic model for BPA.
Effect of initial concentration of Cr(VI) and BPA. Figure 8 (a,b) shows the effect of initial concentration of studied pollutants on the degradation of Cr(VI) ions with respect to the degradation ratio (C/C o ) and various intervals of time (t). Figure 8(a) shows the photodegradation of 25 mg L −1 Cr(VI) ions with a catalyst dosage of 0.1 g L −1 under pH 2 with a wavelength of 254 nm and with 4 lamps intensity (32 W). Figure 8(b) shows the photodegradation of 50 mg L −1 Cr(VI) ions under similar conditions. The concentration of Cr(VI) ions remains same in the absence of photocatalyst. As the initial Cr(VI) concentration increases, there is a decrease in the degradation of Cr(VI) ions from 60 to 45% in 220 min for 25 and 50 mg L −1 of Cr(VI) respectively. This is because there are less electron hole pairs as the concentration of Cr(VI) ions is high. So, there will be a decrease in the degradation process. Figure 8(c,d) shows the effect of initial concentration on the degradation of BPA with respect to the degradation ratio (C/C o ) and at various intervals of time (t).  Effect of solution pH. Figure 9(a,b) shows the effect of solution pH on the degradation of Cr(VI) with respect to the degradation ratio (C/C o ) and various intervals of time (t). Figure 9(a) shows the photodegradation of 10 mg L −1 Cr(VI) ions with a photocatalyst dosage of 0.1 g L −1 at pH 2 with a wavelength of 254 nm and with 4 lamps intensity (32 W). Figure 9(b) shows the photodegradation of 10 mg L −1 Cr(VI) at pH 5.6 under similar conditions. As seen from Fig. 9(a,b), Cr(VI) was well degraded in acidic pH than normal pH (5.64). At pH 2, Cr(VI) ions degraded almost 100% within 140 min. Whereas, Cr(VI) ions were degraded approximately 20 to 30% of the initial concentration under normal pH (5.64) ( Fig. 9(b)). It reveals that in acidic pH, more negatively charged Cr(VI) species in the form of CrO 4 2− were associated with the positive charged surface of photocatalyst via electrostatic attraction which leads to the substantial degradation of Cr(VI) ions in acidic pH 27 . As the pH increases from 2 to 5.64, there is a decrease in the degradation of Cr(VI) ions, which is due to the electrostatic repulsion of negative charged surface and negative charged Cr(VI) species on the photocatalyst surface. Hence, pH 2 was selected for the degradation of Cr(VI) for the remaining parameters. Figure 9(c,d) shows the effect of pH on the degradation of BPA with respect to the degradation ratio (C/C o ) and at various intervals of time (t). Figure 9(c) shows the photodegradation of 10 mg L −1 BPA with a photocatalyst dosage of 0.1 g L −1 under pH 2 with a wavelength of 365 nm and with 4 lamps intensity (32 W). Figure 9(d) shows the photodegradation of 10 mg L −1 BPA under pH 5.64 with similar conditions. After observing Fig. 9(c,d), BPA is well degraded at normal pH than acidic pH. At pH 2, degradation of BPA is less, because at acidic pH, the surface of the photocatalyst and BPA were positively charged, hence, there was repulsive force between them. Therefore, little degradation occurred in acidic pH. In the case of normal pH (5.64), the degradation of BPA is quite good and it degraded upto 90% in 240 min, where the positive charge of BPA and negative charge of catalyst favors the degradation 28 . Hence, normal pH was chosen for the degradation of BPA for the remaining parameters.
Effect of UV wavelength. Figure 10(a,b) shows the effect of UV wavelength on the degradation of Cr(VI) ions with respect to the degradation ratio (C/C o ) and various intervals of time (t). Figure 10(a) shows the photodegradation of 10 mg L −1 Cr(VI) ions with a photocatalyst dosage of 0.1 g L −1 under pH 2 with a wavelength of 254 nm and with 4 lamps intensity (32 W). Figure 10(b) shows the photodegradation of 10 mg L −1 Cr(VI) ions under the wavelength of 365 nm under similar conditions. It was found that both UV wavelengths (254 and 365 nm) have considerable effect on the degradation of Cr(VI) ions, but in the case of 254 nm, Cr(VI) ions were more effectively degraded in a short period of time, as compared to 365 nm. Hence, for all the parameters, 254 nm wavelength was chosen for degradation of Cr(VI) ions. Figure 10(c,d) shows the effect of UV wavelength on the degradation of BPA with respect to the degradation ratio (C/C o ) and various intervals of time (t). Figure 10  have considerable effect on degradation of BPA, but in the case of 365 nm, BPA was more effectively degraded as compared to 254 nm. Hence, for all the parameters, 365 nm was chosen for the degradation of BPA.
Effect of light intensity. Figure 11(a,b) shows the effect of light intensity on the degradation of Cr(VI) ions with respect to the degradation ratio (C/C o ) and various intervals of time (t). Figure 11(a) shows the photodegradation of 10 mg L −1 Cr(VI) ions with a photocatalyst dosage of 0.1 g L −1 under pH 2 with wavelength of 254 nm and with 4 lamps light intensity (32 W). Figure 11(b) shows the photodegradation of 10 mg L −1 Cr(VI) ions under 2 lamps light intensity (16 W) under similar conditions. Both the intensities were found effective for the degradation of Cr(VI) ions with different degradation times. Cr(VI) ions were degraded within 140 min with 32 W light intensity ( Fig. 11(a)), but with 16 W light intensity, degradation of Cr(VI) was slower (Fig. 11(b)). Hence, for the all parameters, 32 W light intensity was chosen for Cr(VI) ions degradation. Figure 11(c,d) shows the effect of light intensity on the degradation of BPA with respect to the degradation ratio (C/C o ) and various intervals of time (t). Figure 11(c) shows the photodegradation of 10 mg L −1 BPA with a photocatalyst dosage of 0.1 g L −1 under pH 5.64 with a wavelength of 365 nm and with 4 lamps intensities (32 W). Figure 11(d) shows the photodegradation of 10 mg L −1 BPA under 2 lamps light intensities (16 W) with similar conditions. It was found that both the intensities helped in degradation of BPA also with degradation times. BPA was degraded 90% within 240 min with 32 W light intensity (Fig. 11(c)), but with 16 W light intensity, degradation of BPA was slower (Fig. 11(d)). Hence, for the all parameters, 32 W light intensity was chosen for BPA degradation. Table 2 shows the comparison between different photocatalysts with the prepared photocatalysts (in this study), which reveals that the photocatalysts (prepared in this study) show the better performance in the degradation of Cr(VI); however, in case of BPA, a comparable removal efficiency was found 26,[29][30][31][32][33][34] .
The absorbance spectra of Cr(VI) and BPA under UV-irradiation are shown in Fig. S6(i & ii). Cr(VI) degradation was monitored by the time-evolution of absorbance spectra. Fig. S6(i) shows the typical absorbance spectra of Cr(VI) solution with irradiation time with all the three photocatalysts. Among all the three photocatalysts, magnetic 3D-TiO 2 @HPGA NC showed the better degradation efficiency within 140 min. As seen from the figure, the absorbing intensity of Cr(VI) decreases with irradiation time, indicating the rapid photodegradation of Cr(VI). The same was also noticed for the degradation of BPA (Fig. S6(ii)). Effect of scavengers and TOC analysis. In photodegradation experiment, the pollutants are oxidized by some reactive species such as holes (h + ), superoxide radicals (O 2

•−
) and hydroxyl radicals ( • OH). The radical trapping experiment was performed to identify the reactive species involved in the photodegradation of BPA by m3D-T-HPGA NC with various kinds of radical scavengers such as ammonium oxalate (AO), benzoquinone (BQ) and isopropyl alcohol (IPA) and the obtained results are depicted in Fig. 12(a). In this experiment, IPA (1 mM) was added to scavenge • OH and 89% degradation was observed when compared to no scavengers (90%).   This revealed that • OH did not involve in the degradation reaction. Furthermore, the addition of BQ (1 mM) and AO (1 mM) scavenged O 2 •− and h + . About 8% and 72% of BPA degradation was observed by BQ and AO, respectively, when compared with no scavenger. The results revealed that O 2 •− and h + are the important reactive species responsible for the degradation of BPA. Further, it could be concluded from the results that O 2 •− and h + are the major and minor species for the photocatalytic degradation of BPA by m3D-T-HPGA NC.
In addition, total organic carbon (TOC) analysis was performed to investigate the mineralization of photo-degraded BPA solution using m3D-T and m3D-T-HPGA NC. The mineralization of BPA was calculated to be 24 and 67% respectively on m-3D-T and m3D-T-HPGA NC. (Fig. 12b). A 4-fold enhancement was achieved using m3D-T-HPGA NC, clearly indicating the efficiency of m3D-T-HPGA NC for mineralization of BPA solution.
Photostability and reusability studies. The photostability and reusability of photocatalyst are of vital importance for its commercial applications. The photostability and reusability features of m3D-T-HPGA NC were evaluated by the photocatalytic degradation of Cr(VI) and BPA under illumination for three cyclic runs. In every cycle, the recovered material was washed, centrifuged and annealed at 60 °C for 6 h. Then the recovered material was weighed again to add the lost amount and used for the next cycle. As shown in Fig. 13(a), the photodegradation efficiency of m3D-T-HPGA NC for the degradation of Cr(VI) and BPA almost remained same even after three cycles, which shows that the prepared m3D-T-HPGA NC is reusable. Figure 13(b) depicts the XRD patterns of fresh and used m3D-T-HPGA NC. From the results, it can be seen that the peak positions and crystal structure of used m3D-T-HPGA NC are similar to the fresh m3D-T-HPGA NC. This inferred that m3D-T-HPGA NC is stable under degradation experiment and it can be used over a long period of time.

Conclusions
Three photocatalyst materials, viz. b-3D-T, m3D-T and m3D-T-HPGA NCs were successfully synthesized by both hydrothermal and solvothermal processes. The prepared nanocomposite (m3D-T-HPGA NCs) showed 100% and 90% degradation of Cr(VI) and BPA, respectively under UV illumination. Present study reports the performance of prepared nanocomposites by varying different parameters such as, photocatalyst dosage, initial  pollutant concentration, solution pH, light wavelength and light intensity. Compared to b-3D-T and m3D-T nanocomposites, m3D-T-HPGA shows the better degradation over Cr(VI) and BPA pollutants. At pH 2 and 5.64, Cr(VI) and BPA pollutants were degraded well with the wavelength of 254 (Cr(VI)) and 365 (BPA) nm and with 32 W light intensity. Overall, the prepared photocatalyst nanocomposite shows the better performance towards the reduction of Cr(VI) and BPA. Decorating magnetic particles onto the 3D-TiO 2 is an advantage which is useful for easy recovery of material from the treated water.
Synthesis of 3D-TiO 2 . The 3D-TiO 2 nanoparticles were synthesized by using hydrothermal method 13 . In brief, titanium tetrachloride (TiCl 4 ) was added drop-wise to the deionized water in an ice bath with vigorous stirring to obtain 50 wt.% of TiCl 4 aqueous solution. Simultaneously, 4 ml of tetrabutyl titanate (TBT) was added drop wise to 30 ml of toluene in an ice bath under vigorous stirring. Subsequently, 50 wt.% obtained TiCl 4 solution was added drop-wise to the TBT solution under stirring for 1 h. After 1 h, the obtained white precipitate was then transferred to the stainless-steel Teflon autoclave and kept for 24 h at 150 °C. After 24 h, the obtained material was centrifuged and washed with DI water and ethanol for several times. Finally, the material was kept for drying in vacuum oven at 50 °C and denoted as bare 3D-TiO 2 (b-3D-T).

Synthesis of magnetic 3D-TiO 2 .
Magnetic 3D-TiO 2 nanocomposite was prepared by using solvothermal process. In brief, 50 ml of ethylene glycol containing 2 mmol of iron chloride and 1 mmol of cobalt chloride was kept for ultrasonication for 30 min. Then, 0.2 g of 3D-TiO 2 nanoparticles were added to the iron-cobalt solution under magnetic stirring for 30 min. Later, 3.6 g of sodium acetate and 1.0 g of polyethylene glycol were added to the mixture and stirred for 30 more min. The mixture was then transferred to stainless-steel Teflon autoclave and kept at 200 °C for 12 h. Finally, the mixture was centrifuged, washed several times with DI water and ethanol, and then kept for drying at 50 °C, and denoted as magnetic 3D-TiO 2 (3D-TiO 2 @CoFe 2 O 4 ) (m3D-T).
Synthesis of magnetic 3D-TiO 2 @HPGA nanocomposite. Graphene oxide (GO) was synthesized by modified Hummer's method. Hierarchical porous graphene aerogel (HPGA) was prepared by the method reported in the literature 21 using GO as precursor material. The magnetic 3D-TiO 2 @HPGA nanocomposite was prepared by using solvothermal process. In brief, 50 ml of ethylene glycol containing 0.5 g of HPGA, 2 mmol of iron chloride and 1 mmol of cobalt chloride was kept for ultrasonication for 1 h. Then 0.2 g of 3D-TiO 2 nanoparticles were added to the iron/cobalt solution under magnetic stirring for 30 min. Later, 3.6 g of sodium acetate and 1.0 g of polyethylene glycol were added to the mixture and stirred for 30 more min. Then the mixture was transferred to stainless-steel Teflon autoclave and kept at 200 °C for 12 h. Finally, the mixture was centrifuged, and washed several times with DI water and ethanol, and then kept for drying at 50 °C, and denoted as magnetic 3D-TiO 2 @HPGA (m3D-T-HPGA NC).
Photocatalytic studies. Heber compact multi wavelength and multilamp photo reactor (Model: HML-COMPACT-LP-MP88) was used in this study for the degradation of Cr(VI) and bisphenol A (BPA). The photocatalytic experiments were done under UV light irradiation. Photoreactor consists of 4 sets of 8 W low pressure mercury vapor lamps (Sankyo denki, Japan) with the maximum light intensity of 254 nm and 365 nm. Photocatalytic experiments were carried out with a known amount (5 mg) of prepared nanophotocatalyst suspended in 50 ml of Cr(VI) and BPA aqueous solution with the concentration of 10 mg L −1 under stirring. Before starting the photodegradation experiments, the mixtures were stirred under dark for 30 min, to obtain the adsorption-desorption equilibrium. Five different parameters were studied in this study, viz. photocatalyst dosage (0.1 and 0.2 g L −1 ), varying pH (2 and 5.6), initial pollutant concentration (10, 25 and 50 mg L −1 ), lamp intensity (16 and 32 W) and wavelengths (254 and 365 nm). Each parameter was studied by keeping other parameters constant. At certain time intervals, 3 ml solution was taken out from the reactor and filtered using 0.45 µm filter. The filtrate was then analyzed by using UV-Visible spectrophotometer at 200-800 nm wavelength. The concentration of Cr(VI) was measured by spectrophotometric method using 1,5-diphenylcarazide as color reagent at 540 nm. The concentration of BPA was measured at 276 nm.
Photoelectrochemical measurements. The photoelectrochemical properties of as prepared photocatalysts were investigated using CHI660C electrochemical workstation with a conventional three electrode system. Ag/AgCl, Pt-wire were served as a reference and counter electrodes and 0.1 M Na 2 SO 4 aqueous solution was used as an electrolyte. The working electrode was prepared by the following procedure: 5 mg of photocatalyst was ground with 20 µL of deionized water and 10 µL of Triton X-100 to prepare slurry. Then the slurry was coated on the conductive side of fluorine doped tin oxide (FTO) plate with an active surface area of about 0.5 × 0.5 cm 2 and then dried in hot air oven at 100 °C for about 6 h.

Data Availability
All data generated or analyzed during this study are included in this published article (and its Supplementary  Information files).