Characterization of TiO2@rGO
The hydrothermal method was utilized to synthesize the composite powders, as shown in Fig. 1a. Use SEM to examine the microstructure of synthesized TiO2 nanoparticles and TiO2@rGO composite powders, as illustrated in Figs. 1b, c. Although the images of the two samples were almost similar, the TiO2@rGO had a pinnate two-dimensional structure, showing the difference with TiO2 nanoparticles. The morphology analysis also clearly showed that the pinnate two-dimensional structure observed in TiO2@rGO was due to the present of rGO. This indicated the adhesion of TiO2 nanoparticles on rGO, and the successful synthesis of the TiO2@rGO, which was conducive to the transfer of photogenerated carriers in the photocatalytic reaction. Previous studies indicated that the migration of photogenerated carriers was crucial in the photocatalytic production of ROS[33].
The XRD patterns of GO, synthetic TiO2 and synthetic TiO2@rGO were illustrated in Fig. 1d. The location of the observed peaks was consistent with previous reports. The peak at 2θ = 11.25°of GO corresponded to the (001) crystal plane[34]. About the TiO2, the peak at 25.28° corresponded to the (101) crystal surface of anatase phase TiO2, indicating that the prepared TiO2 nanoparticle was anatase phase. The specific diffraction peaks of TiO2 at 2θ = 37.91°, 47.86°, 53.75° and 62.68° could also be attributed to the (004), (200), (211) and (204) reflections of TiO2 nanocrystalline surfaces (JCPDS No. 21-1272), respectively[35, 36]. This indicated that the TiO2 nanoparticle had been successfully synthesized. For the TiO2@rGO, the characteristic reflection peaks corresponding to each crystal surface of anatase phase TiO2 were observed, but no specific reflection of GO was observed, indicating that all GO had been successfully reduced to rGO in the hydrothermal reaction process. The characteristic peak of rGO was not observed, which might have been because its characteristic peak at about 27° was close to the (101) crystal peak of TiO2. The functional groups of GO, TiO2 and TiO2@rGO samples were analysed by FTIR, as illustrated in Fig. 1e. The 3450 cm− 1 peaks in all spectra indicated C-OH group's O-H stretching vibration. The peak at 1627 cm− 1 observed in GO was attributed to the C = C skeleton vibration, while the 1400 cm− 1 and 1087 cm− 1 peaks corresponded to the C-OH and C-O-C vibrations, respectively[36, 37]. Absorption peaks at 500–800 cm− 1 were observed in the TiO2 sample, which resulted from the Ti-O-Ti stretching vibration. For TiO2@rGO sample, a peak caused by the skeleton vibration of rGO was observed at 1627 cm− 1. The disappearance of the peak corresponding to C-O-C in rGO at 1087 cm− 1 indicated that GO was successfully reduced to rGO.
The vibration modes of the three samples were studied using Raman spectroscopy, as illustrated in Fig. 1f. Two well-defined peaks were observed in the GO sample at 1349 cm− 1 and 1595 cm− 1, which were attributed to the D and G bands, respectively[36]. The D band, which was a typical characteristic of sp3 defects in carbon, and the G band, which furnished valuable insights into the in-plane vibrations of sp2-bonded carbon atoms of graphene, were both discernible features in the sample[38]. Some oxygen-containing functional groups combined at the boundary of the graphite layer, resulting in a large number of defects C (sp3) and impurities, which destroyed the crystal structure of graphite. About the TiO2 samples, it showed the peaks of typical anatase phases at 147 (Eg(1)), 391 (B1g), 512 (A1g) and 633 (Eg(2)) cm− 1, respectively. The TiO2@rGO sample displayed the characteristic peaks of graphene, with the D (1349 cm− 1) and G (1595 cm− 1) peaks being detected, respectively. Strength ratio (ID/IG) of 1.21 was much higher than that of GO (0.85), indicating that GO had been successfully converted to rGO[39]. The peaks of TiO2 (141, 386, 506 and 634 cm− 1) were also existent, indicating the synthesis of TiO2@rGO composite powders.
XPS was conducted to analyze the composition of TiO2 and TiO2@rGO samples. Analysis of the obtained spectra revealed the existence of O, Ti, and C elements, and the outcomes were presented in Fig. 1g. The high-resolution XPS spectra of C, Ti and O were analyzed and displayed in Figs. 1h-j. C 1s binding energy peaks measured in the TiO2 sample were at 284.8 eV (C-C), 286.4 eV (C-O) and 288.3 eV (C = O), while that of the TiO2@rGO was at 284.6 eV, 286.5 eV, and 288.9 eV[36]. However, an additional binding energy peak located at 285.6 eV was observed in the TiO2@rGO sample. Combined with the above characterization experiments, it was judged to be a C-O-Ti bond, indicating that TiO2 and rGO had chemically bonded, which was advantageous for facilitating the transfer of photogenerated charge carriers. The binding energies of Ti 2p1/2 and Ti 2p3/2 in TiO2 were 464.4 eV and 458.6 eV, respectively, while those of the TiO2@rGO moved to 465.2 eV and 459.7 eV, respectively. It indicated that there was electronic transfer from TiO2 to rGO, and led to an increase in TiO2 binding energy. As shown in Fig. 1j, the binding energies of O 1s of the TiO2 sample were located at 529.9 eV (Ti-O bond) and 530.7 eV (C-O bond), while the binding energy peaks of the TiO2@rGO sample moved to 530.7 eV (Ti-O bond) and 531.2 eV (C-O). The C = O bond at 533.2 eV was also observed. The results of XPS analysis confirmed a bonding connection was established between TiO2 and rGO, which promoted electron transfer.
The photocatalytic activity of the above two samples was investigated and characterized. The light absorption characteristics of them were measured using UV-vis spectroscopy, as illustrated in Fig. 2a. The absorption band edge of the TiO2 sample was around 390 nm, similar to TiO2@rGO, which was consistent with previously reported results[40]. rGO did not significantly broaden the optical response range of TiO2. Subsequently, the photocatalytic activity was characterized under UV light at 365 nm. The effect of rGO on photogenerated electron-hole pair separation was demonstrated by transient photocurrent response and Electrochemical Impedance Spectroscopy (EIS) measurement[41]. An equivalent circuit diagram was obtained through EIS measurement and was shown in Fig. 2b, with the measured results presented in Fig. 2c. The diameter of the Nyquist circle was positively correlated with the impedance, which reflected the relative size of the electrochemical impedance[41]. The TiO2@rGO sample exhibited a shorter Nyquist circle diameter than the TiO2 sample, indicating much smaller interfacial charge transfer resistance and better charge transfer efficiency.
The efficiency of photogenerated carriers’ separation and conversion in the photocatalyst was studied through a series of experiments. Photoluminescence (PL) is an effective technique to evaluate the efficiency of photoluminescence electron-hole separation[41, 42]. When the photocatalyst absorbs light energy to produce photogenerated carriers, the subsequent recombination of these charge carriers results in the emission of light. Generally, the intensity of photoluminescence is directly proportional to the degree of separation of photogenerated electron-hole pairs. Higher carrier separation efficiency causes lower PL intensity, resulting in more efficient photocatalytic behavior. The PL spectra of the two samples at atmospheric temperature under a 365 nm UV light were presented in Fig. 2d, featuring a broad emission peak centered at 435 nm. The higher PL strength of the TiO2 indicated that the holes and electrons in TiO2 were easier to recombine, confirming that the addition of rGO was conducive to the inhibition of carrier recombination.
Transient photocurrent responses of interfacial charge separation were studied by collecting multiple switching cycles under a 365 nm UV light, as shown in Fig. 2e. As is known, a high separation rate of photon-producing carrier supports photocatalytic efficiency. The stronger index of photocurrent intensity, the higher the photocatalytic efficiency and activity. As illustrated in Fig. 2f, the photocurrent density increased when the power was turned on and the illumination started, while the current density decreased when dark. The superior photocurrent intensity observed in TiO2@rGO compared to that of TiO2 suggested that the rGO effectively enhanced the separation efficiency of photogenerated carriers. In addition, as shown in Figs. 2g-i, ESR was used to detect the ROS produced by the TiO2@rGO under photocatalytic conditions. The signal of ·OH was the strongest, followed by that of ·O2−, while there was almost no signal of 1O2[41]. It indicated that ·OH was the main product of the TiO2@rGO photocatalysis, with a small amount of ·O2−, and both of these were the main factors contributing to the photocatalytic antibacterial activity.
Characteristic of scaffold
The PLLA powders, the prepared PLLA/TiO2 and PLLA/TiO2@rGO composite powders were processed into scaffolds with personalized porous structure using the SLS technique and labeled respectively as PLLA, PT and PTG, as shown in Fig. 3a. The designed porous support was shown in Fig. 3b. The scaffold was cylindrical (diameter 12 mm, height 12.5 mm) and had a number of through-holes distributed on the upper front, as well as a large number of lateral holes on the side. These features were conducive to nutrient delivery and cell proliferation. FTIR and XRD were utilized to analyze the scaffold's composition, and the corresponding results were illustrated in Figs. 3c, d, respectively. In the FTIR spectra of PTG exist a peak at 500–800 cm− 1 indicating the presence of TiO2. Similarly, in the XRD patterns, the peak observed at 2θ = 25.3° for the PT and PTG samples indicated the presence of TiO2 in the scaffolds. The results of FTIR and XRD confirmed that the scaffolds were successfully prepared.
The mechanical properties of the PLLA, PT and PTG scaffolds were evaluated by tensile and compression tests, and the relevant results are shown in Figs. 3e, f and Figs. 3g, h. The scaffolds containing TiO2 and TiO2@rGO exhibited enhanced tensile strength and modulus. Specifically, the PTG scaffold demonstrated a notable increase in tensile strength from 474 MPa to 640 MPa compared to the PLLA scaffold, while the modulus was only slightly elevated. Moreover, the compressive strength and modulus of the scaffolds containing TiO2 and TiO2@rGO were also enhanced. Specifically, the PTG scaffold demonstrated an increase in compressive strength from 130 MPa to 230 MPa and an increase in compressive modulus from 14.73 MPa to 18.77 MPa, resulting in a 76.9% and 27.4% improvement, respectively, compared to the PLLA scaffold. Incorporating TiO2 and TiO2@rGO as reinforcement phases resulted in an improvement of the mechanical properties of the PLLA scaffold.
Photocatalytic antibacterial effect in vitro
To investigate the antibacterial ability of the PLLA, PT and PTG scaffolds, antibacterial experiments were conducted in vitro. The impact of the scaffolds on the proliferation of E. coli and S. aureus were observed by fluctuations in bacterial population on the petri dish, as shown in Figs. 5a, b. Taking the PLLA scaffold in dark as the negative control group, the number of bacterial colonies on the PT scaffold decreased slightly, and that on the PTG scaffold decreased significantly. This indicated that TiO2 had a photocatalytic antibacterial effect, and the addition of rGO effectively improved the photocatalytic antibacterial effect of TiO2. As biofilm formation is a crucial factor in implant-associated infections, we examined the scaffolds' ability to eliminate biofilms. The live/dead staining experiment of bacteria biofilm treated by photocatalysis showed that UV irradiation of the PTG scaffold had a strong destructive effect on the bacterial biofilm, as illustrated in Figs. 5c, d. However, only a small amount of the bacterial biofilm was destroyed under UV irradiation alone. The bacterial counts of each group in the plate counting experiments were counted and the survival rate was calculated, as shown in Figs. 5e, f. Under dark conditions, both the PT and PTG scaffolds had a weak inhibitory effect on bacteria due to the presence of TiO2 and rGO. Compared to dark conditions, UV irradiation led to a slight decrease in the survival rate of both bacterial strains. Upon exposure to UV irradiation, the PTG scaffold exhibited a considerable reduction in bacterial survival rate, with E. coli and S. aureus displaying survival rates of 40% and 29%, respectively, compared to survival rates of 62% and 45% for E. coli and S. aureus in the PT scaffold. This indicated that rGO enhanced the photocatalytic activity of TiO2 in producing ROS, thereby enhancing the antibacterial effect.
The morphological changes of the bacteria on the PLLA, PT and PTG scaffolds without or under UV irradiation were studied by SEM characterization, as shown in Figs. 5g, h. The untreated E. coli and S. aureus in the control group were along with smooth and intact cell walls. In all the unilluminated groups, the morphologies of the bacteria remained almost unchanged. The PT scaffold was irradiated by UV light, and a few bacteria were damaged, and the surface became wrinkled[44]. However, the PTG scaffold experienced serious damage to bacteria after UV irradiation, resulting in unclear bacterial boundaries, different degrees of surface depression and even cell wall damage. These results demonstrated that the PTG scaffold could destroy existing biofilms and prevented biofilm formation under UV irradiation, as shown in Fig. 6. Previous studies indicated that TiO2 could generate electron-hole pairs under UV irradiation. Due to the low electrochemical impedance of rGO, electrons transfer from the TiO2 surface to rGO, and reacted with O2 adsorbed on the rGO surface to produce •O2−. At the same time, photo-generated holes reacted with H2O to generate •OH. As ROS including •O2− and •OH attacked bacteria. They destroyed the structure of bacterial biofilms, and furthermore, damaged bacterial genetic material.
Cytocompatibility
Cell assays were conducted to assess the biocompatibility of the PTG scaffold, with the PLLA scaffold serving as the control. Figure 7a depicted SEM images of hBMSCs cultured on the scaffolds for 1, 3 and 7 days. Within 1 day, hBMSCs on the PLLA scaffold exhibited a slim spindle shape with distinct separation from one another. Some filamentous pseudopodia extensions were observed on the PTG scaffold. After 3 d, cells coverage area increased and an irregular polygonal shape emerged as the filamentous extensions intermingled. After 7 d, hBMSCs had grown and extended on the PTG scaffold to form a cell layer. It suggested that the PTG scaffold had a better cytocompatibility, indicating that the addition of rGO and TiO2 in PLLA could promote cell adhesion.
Cell viability for both PLLA and PTG scaffolds was assessed using fluorescent staining techniques to distinguish living (green) and dead (red) cells, as illustrated in Fig. 7b. Obviously, living cells spread well and increased over time, as the filamentous pseudopodia grew, and there were few dead cells in the scaffolds[45]. Additionally, consistent with the cell adhesion experiment, the PTG scaffold showed higher cell viability than the PLLA scaffold at each sampling time. It suggested that TiO2@rGO could promote cell proliferation of PLLA. In addition, the relative cell area, cell density and CCK-8 result of each group were statistically analyzed and calculated, as shown in Figs. 7c-e. The superior ability of the PTG scaffold to promote hBMSCs growth may be attributed to several factors. Firstly, the incorporation of TiO2@rGO improved the scaffold's hydrophilicity, which facilitated cell adhesion and proliferation. Additionally, the rough and large surface area of the TiO2@rGO offered more sites for cell adhesion, and the presence of Ti element was beneficial for stimulating cell growth.