The XRD patterns of TiO2/graphene composites before and after hydrogenation process are shown in Fig. 1(A). The XRD patterns of the TiO2/graphene composites before hydrogenation process show only some diffuse peaks, indicating that the TiO2/graphene composites before hydrogenation are amorphous [20]. After the hydrogenation process, the peaks occur at 25.28° (101), 37.80° (004), 48.05° (200), 53.89° (105), 55.07° (211), 62.69° (204), 68.93° (116), 70.31° (220), and 75.03°(215) (Fig. 1A(b)), corresponding to the diffractions of anatase TiO2 (JCPDS 21-1272) [18, 21]. It was indicated that the sample after surface hydrogenation is crystallized from amorphous to anatase structure, and the average crystal size is approximately 21 nm calculated by Scherrer formula, in agreement with TEM observation.
The structure of black TiO2/graphene composites can also be characterized by Raman spectra. The Raman spectra of BTG with different graphene contents are shown in Fig. 1(B). The three bands at around 1365 cm-1 (D band), 1580 cm-1 (G band) and 2700 cm-1 (2D band) are corresponding to graphene [22]. For all the BTG with different graphene contents, the Raman peaks occur at around 156 cm-1 (Eg(1)), 406 cm-1 (B1g(1)), 523 cm-1 (A1g + B1g(2)), and 646 cm-1 (Eg(2)), which match with the characteristic peaks of anatase TiO2 [23]. Compared with the characteristic peaks of anatase TiO2, the Eg(1) mode shift from 144 cm-1 of bare bulk TiO2 to 156 cm-1 of BTG synthesized. The shift toward high frequency indicates the ultra-dispersed characteristics of TiO2 nanoparticles and their combination with graphene [24]. And the disappearance of the graphene 2D band in BTG may be attributed to the composite of graphene and TiO2 [11]. From the Raman analysis, the characteristic peaks of TiO2 and graphene appeared in the spectra of BTG, indicating that the black TiO2/graphene composites were successfully synthesized.
Furthermore, the elemental composition of the black TiO2/graphene composites was investigated by XPS. In Fig. 2(A), the characteristic peaks of C 1s, Ti 2p, and O 1s are present at 284.6, 457.8, and 529.7 eV respectively. In the Ti 2p XPS spectrum (Fig. 2(B)), the Ti 2p3/2 and Ti 2p1/2 of TiO2 are revealed at 457.8 and 463.3 eV. The Ti 2p3/2 peak of BTG shifts from 458.6 eV to a lower binding energy corresponds to the presence of a high Ti3+ concentration [25-27]. All Ti 2p spectra are symmetrical on the low energy side, indicating that TiO2 is not doped with carbon [17]. The curve fit of C 1s spectra of BTG is shown in Fig. 2(C). The peak at 284.5 eV is ascribed to the C=C/C-C bond, indicating the presence of graphene. The weak peak at 286.5 confirms the presence of the C-O bond [28]. In addition, there is no Ti-C peak observed in both Fig. 2(B) and (C), which confirms that graphene does not exist as a dopant in the BTG composites. The curve fit of O 1s spectra of BTG is shown in Fig. 2(D). The peak at 529.6 and 532.2 eV are ascribed to Ti-O and C-OH bonds. In both Fig. 2(C) and (D), the appearance of C-O and C-OH bonds indicate the existence of bond between carbon and oxygen atoms in the BTG composites. This phenomenon may be attributed to the oxidation of graphene since TiO2 is a well-known catalyst.
The morphology, structure and element distribution of the black TiO2/graphene composites were examined by TEM. The TEM observation in Fig. 3(A) shows that the TiO2 nanoparticles are well dispersed on the graphene sheets, and 10–50 nm in size.
The selected area electron diffraction (SAED) pattern is shown in the inset of Fig. 3(A). The (101), (004), (200), (204) and (105) diffraction rings are detected in the SAED pattern indicates that TiO2 in the composites is anatase structure [18]. The SAED result is consistent with the XRD characterization, which indicates that the TiO2 in the composite is anatase structure with excellent photocatalytic activity. Fig 3(B) shows the HRTEM image. It can be seen that the size of individual TiO2 nanocrystal was approximately 15 nm in diameter. There was a disordered surface layer surrounding the TiO2 nanocrystal, as shown by the dotted red circle in Fig 3(B). The thickness of the disordered layer is about 1 nm, which is consistent with the black TiO2 reported by Chen et al [17]. The disordered layer surrounding TiO2 nanocrystal is created by hydrogenation, which causes a significant color change and enhancement of visible light photocatalytic activity. The schematic diagram of sample color change (from blue to black) after hydrogenation is shown in Scheme 1. The inset in Fig 3(B) shows that the interplanar spacing is 3.58 Å, corresponding to the (101) plane of anatase TiO2. The Energy Dispersive X-ray (EDX) elemental mappings of Ti, C, O taken from the STEM image of Fig 3(C) are given in Fig 3(D-F), respectively. It can be seen from the figures that the Ti and O elements are uniformly aggregated and dispersed on the C element of graphene, which is consistent with the TEM result of Fig. 3 (A). The results further demonstrate the successful assemblies of TiO2 on graphene in the black TiO2/graphene composites.
Fig. 4(A) shows the UV-vis diffuse reflectance spectra (DRS) of the black TiO2/graphene composites synthesized. The presence of graphene significantly improves the visible light absorption of the black TiO2/graphene composites. The visible light absorption intensity of the black TiO2/graphene composites is enhanced with increasing graphene content. In order to characterize the band gaps of the black TiO2/graphene composites, the Kubelka-Munk function (F(R∞)·E)1/n versus the energy of light (E=hv) is shown In Fig. 4(B) [29, 30]. For an indirect transition of anatase TiO2, n=2 will give the best linear fit. As the graphene content increases from 1% to 15%, the band gaps are estimated roughly to decrease from 3.05 to 2.94 eV. It is well known that the band gap energy of anatase TiO2 is 3.2 eV. The band gap around 3 eV of black TiO2/graphene composites is lower than that of anatase TiO2, which could be attributed to the self-doping of Ti3+. Moreover, the black TiO2/graphene composites have enhanced light absorption in the range of visible light, which is consistent with the darker sample color with increasing graphene content. The results suggest that both graphene combination and self-doping of Ti3+ play a crucial role in the photocatalytic activity of the black TiO2/graphene composites. The photocatalytic activity of the samples was shown in Fig. 4(C). Fig. 4(C) illustrates the normalized MB concentration in the degradation solution as a function of visible light irradiation time. After a visible light irradiation time of 60 min, 95%, 98%, 99% and 96% MB are decomposed in the presence of the BTG-1, BTG-5, BTG-10 and BTG-15, respectively. A comparative experiment without catalyst during visible light irradiation exhibits only 15% MB decomposition. The photocatalytic process follows first-order kinetics, c = c0exp(-kt), where c0 and c are the MB concentration before and after visible light irradiation, respectively [31]. The k value in the formula represents the photocatalytic reaction rate. Through fitting calculation, the photocatalytic reaction rates k for BTG-1, BTG-5, BTG-10 and BTG-15 are determined to be 2.88, 3.61, 3.99 and 3.23 h-1, respectively. The BTG-10 exhibits the highest photocatalytic activity. To determine the recyclability of the black TiO2/graphene composites, the BTG-10 is recycled under several visible light irradiation cycles. As shown in Fig. 4(D), after 5 cycles, the degradation rate of MB still reaches 90% in the presence of the BTG-10, indicating that the catalyst has good stability.
Since BTG-10 has the best catalytic activity in the TiO2/graphene composites, BTG-10 was selected for valence band (VB) XPS analysis in Fig. 5(A). The VB of BTG-10 is located at 2.68 eV, which is lower than the commonly used TiO2 (3.0 eV) [26]. The insets in Fig. 5(A) show the energy band diagrams. According to the energy band model, the conductance band (CB) can be calculated by CB = VB – Eg, where Eg represents the energy of the band gaps. The Eg of BTG-10 is estimated to be 3.0 eV by Fig. 4(B), so the CB of BTG-10 is calculated as -0.32 eV. The results suggest that the disordered surface layer surrounding the TiO2 nanocrystal introduced by hydrogenation can upshift both the VB and CB edge of the TiO2/graphene composites synthesized. According to the energy band analyses, the VB of the BTG-10 is higher than the O2/H2O and CB is H+/H2 potential, suggesting that the black TiO2/graphene composites synthesized have attractive potential for applications in environmental and energy issues.
The photocatalytic mechanism is also proposed as shown in Fig. 5(B). The disordered surface layer introduced by hydrogenation narrows the band gap of the black TiO2/graphene composites, which improves the optical absorption properties [32]. Consequently, the electrons in the VB can easily transition to the CB of TiO2 under visible light irradiation. It is well known that the graphene which nano-sized black TiO2 attached to has good electrical conductivity [33]. Therefore, electrons will be transferred to graphene instead of CB, which is conducive to reducing the opportunities of electron-hole recombination and enhancing photocatalytic activity. However, the graphene itself has no photocatalytic activity, and excessive graphene will hinder the absorption of photons by TiO2. This hinder effect is the reason why BTG-10 has a higher catalytic activity than BTG-15.