A self-organized TiO₂ nanotube (NT) has been widely considered one of the most promising photocatalysts because of its functional properties [1,2], which include high photocatalytic activity, photo-stability and non-toxicity [3,4]. Moreover, the TiO₂ nano-tubular architectures can be precisely controlled by the titanium anodizing process. The TiO₂ nanotubes also exhibit a high specific surface area and anisotropic properties [5-7]. However, their utility was restricted by a relatively large band gap (~3.2 eV for anatase) and fast recombination of the photo-generated electron-hole pairs. Especially, in regards to recombination of photo-generated carries, electrons and holes generated by light absorption may be recombined before they participate in catalytic reactions, and the recombination is detrimental to the photocatalytic efficiency. Therefore, to maximize the photocatalytic efficiency, the recombination of electrons and holes must be delayed by separation of photo-induced electrons and holes [8].

Recently, graphene has attracted research interest for its excellent transparency [9,10], high mobility of charge carriers [11,12] and electrical conductivity [13,14]. These properties of graphene will be a great help in improving the photocatalytic efficiency of TiO₂. In this study, to improve the photocatalytic activity of titania nanotube catalyst, we attempted to synthesize a graphene/Ag/TiO₂ NT photocatalyst. A polymer compound containing CH₃COOAg and poly(L-lactide) (PLLA) was utilized for deposition of the Ag nanoparticles (NPs) on the TiO₂ nanotubes. The silver nanoparticles were precipitated by reducing the silver ions in CH₃COOAg/PLLA during the annealing process. Graphene deposition on the Ag/TiO₂ NTs was achieved using an electrophoretic deposition (EPD) process. The surface characteristics of the graphene/Ag/TiO₂ NT photocatalyst were investigated using field potenemission scanning electron microscopy (FE-SEM) and Fourier transform infrared spectroscopy (FTIR). Photocatalytic activity was evaluated using dye decomposition rates.

Commercial-grade (99.6 wt%) titanium plates were used for anodization. The Ti plate was cleaned in acetone and ethanol, and then rinsed with deionized water for 10 min. The anodic TiO₂ nanotubular film was synthesized by electrochemical anodization at a constant voltage of 30 V in 0.5 wt% NH₄F and 5 wt% H₂O containing ethylene glycol (EG) solution. The crystal structure of the titania nanotubes fabricated by anodizing has been identified as an amorphous TiO₂ phase [15]. For crystallization of anodic titania nanotubes, additional annealing treatment was needed.

Silver acetate (AgC₂H₃O₂, 99.0%; Kojoma Chemical Co.) and PLLA (Fluka) were used to synthesize the Ag NPs on the titania nanotubes. Silver acetate and PLLA were mixed in the weight ratio of 2:1, and the compound was dissolved in CH₂Cl₂. The solution was dropped on the TiO₂ nanotube surfaces under ambient conditions, and then heated in a furnace. To fabricate the crystalline TiO₂ nanotubes with Ag NPs, a two-step annealing process was performed. For precipitation of Ag NPs, a titania-nanotubular layer coated with the precursor compound containing AgC₂H₃O₂/PLLA, was first annealed at 380℃ for 1 h, and then a second annealing was performed at 500℃ for crystallization of amorphous TiO₂ nanotubes.

The graphene deposition on the Ag/TiO₂ nanotubes was performed using an electrophoretic deposition (EPD) process. Graphene oxide (GO) was purchased from NanoGetter Co. (Seongnam, Korea). The GO was chemically reduced to graphene by reaction with hydrazine [16]. For this purpose, 40 mL of aqueous GO solution was mixed with 0.04 mL of 80% hydrazine aqueous solution and 0.28 mL of 30% aqueous ammonia in a conical flask. To produce graphene, the mixed solution in the flask was then placed in a water bath at 100℃ for 2 h. The graphene particles synthesized by reduction of GO with hydrazine, exhibited a negative charge on the surface at pH greater than 4 [17], and could be deposited on the surface of a positive electrode under the applied potential. In this study, the electrophoretic deposition of graphene on the surface of Ag/TiO₂ nanotubes was carried out at a constant voltage of 4 V for 2 min, in a two-electrode cell system. The pH of the graphene solution used as electrolyte was kept at 12, and a platinum sheet was used as the counter electrode. A schematic diagram of the EPD process is shown in Fig. 1.

To investigate Ag precipitation in the compound containing CH₃COOAg/PLLA, differential scanning calorimetry (DSC; DSC N 650, Sinco, Twin Lakes, WI, USA) and thermogravimetry (TG; Scinco N-1000, Sinco) were utilized. The crystalline structure of the TiO₂ nanotubes was characterized by X-ray diffraction analysis (D8 discover, Bruker AXS, Karlsruhe, Germany), and their morphologies were observed using FE-SEM (JSM 7600F, JEOL, Tokyo, Japan). The dye decomposition rates for the photocatalytic efficiency of the TiO₂ nanotube were evaluated using aniline blue [18]. In addition, ultraviolet (UV)/Visible spectroscopy (Unicam 8700, Spectronic Camspec Ltd, Leeds, UK) was performed to determine the decomposition rate of the dye in the absorption band of 600 nm.

The morphology of anodic titania nanotubular films can be controlled by various fabrication parameters. The anodic potenetial applied and the electrolyte composition, in particular, clearly affects the nanotube arrangement and surface morphology of TiO₂ nanotubes. Longer tube-length can be obtained with TiO₂ nanotubes fabricated in an ethylene-glycol-mixture solution than fabricated in acidic electrolytes. Fig. 2 shows the regular nano-tubular arrangement on the self-organized TiO₂ layer synthesized at 30 V in a mixture of EG/0.5 wt% NH₄F/5 wt% H₂O. The anodic TiO₂ layer consisted of nanotubes with an outer diameter of 100 nm. The anodic TiO₂ nanotubes without annealing were identified as being in amorphous phase. Thus, additional annealing treatment was needed for crystallization.

In order to investigate the effect of annealing temperature on the precipitation of Ag NPs, the thermal behavior of the polymer compound containing CH₃COOAg/PLLA (in the weight ratio of 2:1) was analyzed using DSC and TG. The DSC and TG analyses were carried out with 10 mg samples in N₂ gas at a heating rate of 10℃ /min. Fig. 3a shows the thermal reaction of the precursor polymer with temperature, and a large endothermic peak exhibited in the range 320℃-380℃ due to the thermal decomposition of the precursor polymer (PLLA). In this temperature range, the Ag⁺ ions in CH₃COOAg/PLLA may be reduced to silver nanoparticles (Ag^o) since the polymer PLLA acts as a reducing agent [19]. The mass loss in the TG curve can be interpreted in terms of the complex decomposition process related to evolution of CO, CO₂, COOH, and H₂O.

To confirm the precipitation of Ag NPs by reducing Ag⁺ to Ag^o, the polymer containing CH₃COOAg/PLLA was annealed at 380℃ for 1 h, and then the crystalline structure of the compound was examined. The X-ray diffraction (XRD) result of the polymer compound is shown in Fig. 3b. No specific crystalline structure was identified from the XRD pattern of the polymer compound without annealing treatment. However, when the polymer compound was annealed at 380℃ for 1 h, multiple peaks at 2θ = 38.32°, 44.50°, 64.61°, 77.54°, and 81.68° were obvious, and were identified as the (111), (200), (220), (311), and (222) planes of the silver crystalline structure. The XRD result shown in Fig. 3b indicates that the Ag metal particles can be obtained from the precursor containing CH₃COOAg/PLLA, by annealing at 380℃.

In order to precipitate the Ag NPs on the TiO₂ nanotubes, a polymer compound containing CH₃COOAg/PLLA in the ratio of 2:1 was utilized. To reduce the Ag⁺ to Ag^o metal particles in the polymer compound, and to crystallize the amorphous TiO₂ nanotubes, the compound was used to wet the TiO₂ nanotubes, after which a two-step annealing process was performed. The first annealing treatment was at 380℃ for 1 h. Because the PLLA posited with the Ag NPs, and of the graphene layer, were also supplies electrons and acts as reducing agent, the silver ions in the CH₃COOAg were reduced to silver metal particles. During the second annealing at 500℃ for 1 h, the amorphous phase of titania nanotubes was transformed to anatase-type crystalline structure. Moreover, during this annealing process at 380℃, the high heat also removed the residual polymer. Fig. 4a shows the scanning electron microscopy (SEM) image of the TiO₂ nanotubular layer deposited with the Ag NPs. Fig. 4b and c are representative SEM images observed from the top and a cross-section of the sample, respectively. The precipitated silver nanoparticles on the TiO₂ nanotubes, prepared by the wetting and annealing process (shown in Fig. 4b and c), exhibit average particle-size of 10-30 nm. The size of the metal particles can be controlled by adjusting the concentration of the compound polymer. The chemical composition of the surface of the titania nanotubes in Fig. 5a was determined by energy dispersion X-ray spectroscopy (EDS), and the result is shown in Fig. 5b. The elemental signals prove the presence of silver, titanium, and oxygen. The Ag peaks were attributed to the precipitation of Ag NPs on the TiO₂ nanotubes, while the Ti and O peaks were ascribed to the anatasetype TiO₂ nanotubes. X-ray mapping indicated the distribution of the chemical forms of Ti, O, and Ag (Fig. 5c-e). This showed that the precipitated Ag NPs were homogeneously distributed on the surface of the TiO₂ nanotubes. Fig. 6 shows the EDS and X-ray mapping results obtained from the TiO₂ nano-tube cross-section area deposited with the Ag NPs. The result indicates that the Ag NPs were also deposited homogeneously inside the TiO₂ nanotubes, as they were precipitated during the annealing process by reduction of silver ions in the polymer compound.

The deposition of graphene on the Ag/TiO₂ NTs to create the photocatalyst was performed by an electrophoretic deposition process at 4 V in electrolyte containing reduced GO particles at pH 12. For the electrophoretic deposition process, a two-electrode cell was utilized. In this process, negatively charged, reduced graphene-oxide in solution moves toward the positive electrode under an applied DC electric field, therefore, the graphene was gradually deposited onto the Ag/TiO₂ NTs, and the thickness of the graphene layer so deposited, increased with time. Fig. 7 shows the deposited graphene on the Ag/TiO₂ NTs after 2 min of electrophoretic deposition at 4 V. Fig. 7a shows that the surfaces of the Ag/TiO₂ NTs have been covered with graphene sheets, and partially deposited graphene on the Ag/TiO₂ NTs is exhibited in Fig. 7b. Thus, Fig. 7 indicates that the graphene/Ag/TiO₂ hybrid composite for photocatalysis had been synthesized, and that the EPD process was useful for the deposition of graphene on the Ag/TiO₂ NTs.

To investigate phase transformation during annealing at 500℃ for 1 h, the XRD analysis was performed on non-annealed and annealed TiO₂ nanotube layers. Moreover, the crystal structure of the annealed TiO₂ with Ag NPs, annealed TiO₂ deposited with the Ag NPs, and of the graphene layer, were also photocatainvestigated. The resulting XRD patterns are shown in Fig. 8.

For non-annealed TiO₂ nanotubes (Fig. 8a), the XRD pattern exhibited the presence of titanium and anatase peaks, but the anatase peaks were very weakly presented. The titanium peaks originated from the titanium substrate. Thus, the XRD result indicates that most of the crystal structure on non-annealed titania nanotubes was amorphous phase, and that anatase phase was only slightly produced in the anodized TiO₂ nanotubes. However, for annealed TiO₂ nanotubes, as shown in Fig. 8b, the anatase-type crystalline TiO₂ peaks increased. Thus, the annealing treatment changed the amorphous TiO₂ nanotubular film to a crystalline TiO₂ structure. For the annealed Ag/TiO₂ NTs, as shown in Fig. 8c, the anatase peaks increased slightly, and silver peaks observed were weak. However, in the case of the diffraction pattern of annealed graphene/Ag/TiO₂ NTs in Fig. 8d, no distinct changes of the peaks associated with Fig. 8c were observed in the XRD result. This may be attributed to the small graphene content in the TiO₂ hybrid nanotube, as reported elsewhere [20-22].

The FTIR spectra of TiO₂ NTs, Ag/TiO₂ NTs, and graphene/Ag/TiO₂ NTs are shown in Fig. 9. Moreover, for comparison of characteristic features, the spectrum of TiO₂ NTs deposited with graphene (graphene/TiO₂ NTs) is also represented (Fig. 9c). In Fig. 9, the low frequency absorption (below 1000 cm⁻¹) is attributed to Ti–O–Ti vibration [23-26], while the broad absorption from 3000 to 3700 cm⁻¹ is assigned to the O–H stretching vibration of the surface hydroxyl from adsorbed water [26,27]. The absorption peaks at 2930, 2850, and 1455 cm⁻¹ [28] of anodic TiO₂ NT in Fig. 9a, are attributed to the Ti–OH bond [28-30]. For FTIR spectra of Ag/TiO₂ NTs and graphene/Ag/TiO₂ NTs, the absorption bands around 1730 [31] and 1289 cm⁻¹ correspond to carboxylic C-O stretching, and the peaks at 1080 cm⁻¹ are assumed to be affected by the partial presence of functional groups formed by the residual precursor compound containing the AgC₂H₃O₂/PLLA.

For the graphene/Ag/TiO₂ NTs, new absorption bands showed at 1135, 1565, and 1633 cm⁻¹. The absorption band of 1135 cm⁻¹ is attributed to the stretching vibration of C-H and the band at 1565 cm⁻¹ [32,33] is related to the C-C bonds due to skeletal vibration of the graphene sheets. The absorption band around 1633 cm⁻¹ can be attributed to C=C ring stretching [31,34] or to the H–O–H bending band of adsorbed H₂O molecules [34,35]. Therefore, the spectra of the graphene/Ag/TiO₂ NTs showed absorptions associated with graphene (1565 and 1633 cm⁻¹), and these results indicate that the graphene sheets were successfully deposited on the Ag/TiO₂ NT by the electrophoretic deposition process.

In order to investigate photocatalytic efficiency, dye decomposition tests were performed on TiO₂ NTs, Ag/TiO₂ NTs, and graphene/Ag/TiO₂ NTs catalysts. Fig. 10a shows the photocatalytic efficiency of the titania nano-tubular catalysts for dye degradation during irradiation with an Hg lamp. Before the experiment, the starting solution was adjusted to 83 μM aniline blue for decomposition efficiency, and the pH was kept constant at 4.0 throughout the experiment. In order to obtain the actual photocatalytic efficiency of the prepared catalysts, blank tests were carried out, and the photocatalytic efficiencies were compensated. After a dye decomposition test for 5 h, the photocatainvestigatalytic reaction with bare TiO₂ nanotubes caused 26.4% of the dye to be removed after 1 h, 49.3% after 3 h, and 59.5% after 5 h. The corresponding dye removal rates were 33.4%, 58.7%, and 76.1% for the Ag/TiO₂ NTs; and 37.0%, 67.0%, and 85.3% for the graphene/Ag/TiO₂ NTs hybrid catalyst, respectively.

For kinetic analysis of the photocatalytic reaction, the rate constant of the TiO₂ composite catalyst was evaluated. The rate constant for the dye degradation reaction on TiO₂ composites was calculated using eq (1):

where [A] is the concentration of reactant, k is the rate constant, and x is the reaction order. The reaction order for dye decomposition on the TiO₂ NT catalysts has been reported as first order [15]. Therefore, the change of concentration with reaction time can be written as eq (2) and eq (3):

where C_o represents the initial concentration, C_t is the resultant concentration of dye after reaction time (min), and k is the rate constant. The relationship between ln (C_o/C_t) and the photocatalytic reaction time is illustrated in Fig. 10b.

The rate constants for the dye degradation were 2.94 × 10⁻³ min⁻¹ for TiO₂ NTs, 4.62 × 10⁻³ min⁻¹ for Ag/TiO₂ NTs, and 6.26 × 10⁻³ min⁻¹ for the graphene/Ag/TiO₂ NTs composite catalyst, as shown in Fig. 10b. The dye degradation results (Fig. 10a and b), showed that the photocatalytic efficiency was significantly affected by deposition of Ag particles and graphene on the TiO₂ nanotubes. From Fig. 10, the dye destruction efficiency of Ag/TiO₂ NT catalyst was greater than with the TiO₂ NT photocatalyst. Moreover, the graphene/Ag/TiO₂ NT composite catalyst exhibited much higher dye destruction efficiency than did the other TiO₂ photocatalysts.

Regarding the effects of graphene and Ag deposition, Fig. 11 exhibits the photocatalytic degradation mechanism of the graphene/Ag/TiO₂ NTs system. In Fig. 11, the potential of the conduction band of TiO₂ NTs is −4.2 eV [36], and the work functions of Ag [37] and graphene [38] indicate −4.26 and −4.42 eV, respectively. This result suggests that the photo-induced electron transfer from the conduction band of TiO₂ NTs to Ag or graphene was due to a difference in the work function. When light with energy equal to or greater than the band gap of TiO₂ (3.2 eV) is absorbed on the TiO₂ catalyst, photo-excited holes and electrons move to the TiO₂ surface (Process 1), where they participate in catalytic reactions [39-41]. The photo-produced holes H⁺ and OH^• radicals are known as strongly active agents in the dye-degradation process, andOH^• radicals can be generated by oxidation of water.

Moreover, the photo-generated electrons in the conduction band of TiO₂ were consumed by surface absorbed O₂ to form , OH^• and other superoxide radicals.

Due to oxidation of the aniline blue (AB) by OH^• radicals, or to direct oxidation by photo-generated holes, the AB is decomposed.

The photocatalytic decomposition reaction of aniline blue is presented in Fig. 11. In the photocatalytic reaction, many electron–hole pairs could recombine before they reach the surface of the TiO₂. This recombination is detrimental to photocatalytic efficiency. Thus, the photocatalytic rate constant of TiO₂ NTs was much lower due to recombination of the photo-induced charge carriers.

However, if the Ag NPs were homogeneously deposited on the TiO₂ surface, the photo-generated electrons transfer from the conduction band of TiO₂ to the Ag particles (Process 2). Thus, the Ag NPs act as electron traps [42-44], which can effectively decrease the rate of electron-hole recombination, and increase the photocatalytic activity. On the other hand, the photocatalytic efficiency can be decreased by oversized Ag particles on titania surfaces, which act as recombination centers [45]. For the graphene/Ag/TiO₂ NTs composite catalyst, the photo-induced charge recombination could be more effectively suppressed (Process 2, 3, and 4) due to differences in the work function, compared to the other photocatalysts. High efficiency in charge separation leads to increased photocatalytic activity for the dye-degradation reaction. In this work, to create the graphene/Ag/TiO₂ NTs hybrid catalyst, synthesized Ag NPs and graphene sheets were homogeneously deposited onto TiO₂ NTs. This led to a decreased rate of electron-hole recombination and increased photocatalytic activity. Therefore, the graphene/Ag/TiO₂ NTs composite catalyst exhibited much higher dye destruction efficiency due to the synergistic effect of the graphene and the Ag deposition on the titania nanotubes.

To obtain high-performance photocatalytic degradation, a graphene/Ag/TiO₂ NT hybrid catalyst was synthesized, and its fabrication process, surface characteristics, and photocatalytic activity were investigated. The TiO₂ nanotubular photocatalyst was fabricated by anodization, and the Ag NPs on the TiO₂ nanotubes were synthesized by a wetting and annealing process. During annealing in the temperature range of 320℃-380℃, the silver nanoparticles on the titania nanotubes were precipitated by reduction of silver ions in a compound polymer containing AgC₂H₃O₂/PLLA. The deposition of graphene on the Ag/TiO₂ NTs was performed by electrophoretic deposition. The FE-SEM and FTIR results for the graphene/Ag/TiO₂ NTs showed that the Ag NPs and graphene sheets were successfully deposited on the TiO₂ nanotubes by reduction of the compound polymer containing AgC₂H₃O₂/PLLA, and simultaneous electrophoretic deposition.

From the dye degradation result, the photocatalytic efficiency was significantly affected by deposition of Ag NPs and graphene on the TiO₂ catalyst. The Ag-coated TiO₂ NT, compared with bare TiO₂ NT, exhibited a high rate of dye degradation, which is attributed to suppressing the recombination of the hole-electron pairs. The best dye destruction efficiency was obtained with the graphene/Ag/TiO₂ NT photocatalyst. Its high photocatalytic efficiency may be attributed to the synergistic effect of the graphene and Ag deposition onto the TiO₂ NTs.