2019 Volume 60 Issue 8 Pages 1624-1628
Nano-scale noble metallic materials exhibit catalytic activity and can serve as photocatalytic cocatalysts. The effectiveness of nanoporous gold (NPG) as a cocatalyst on TiO2 depends on NPG pore size and composition. Suitably sized pores ensure that NPG can exhibit surface plasmon resonance, which enhances the photocatalytic activity of TiO2 for hydroxyl (OH) radical generation. Moreover, Ag on NPG serves as generation sites for reactive oxygen species, because O2 can be decomposed to atomic oxygen. Owing to these factors, NPG could serve as a high-performance cocatalyst for TiO2. In this study, we utilized the porous property of NPG to fabricate a nanocomposite material of TiO2 and NPG (TiO2/NPG) without annealing or complicated processing. Moreover, we evaluated the photocatalytic activity of the TiO2/NPG composite by measuring the amount of generated OH radicals. Our results suggest that NPG promoted the generation of OH radicals and that the amount of OH radicals increased with particular pore sizes of NPG and amounts of Ag.
Fig. 1 Plane-view SEM images of the NPG surface with pore sizes of (a) 10 nm, (b) 20 nm, (c) 30 nm, (d) 40 nm, and (e) 50 nm, and (f) NPG with TiO2.
TiO2 is frequently used as a catalyst for UV photocatalysis, and its excellent performance in photocatalytic reactions is attributed to the formation of a substantial amount of radicals, such as hydroxyl (OH) radicals and oxygen atoms. An OH radical is one of the radicals that can be generated on the surface of TiO2 by several reaction processes. One of these processes is the oxidation of water molecules by holes, which are generated by the UV excitation of electrons in TiO2. Because OH radicals are highly reactive, they can be used to determine a catalyst’s ability to decompose organic materials in photocatalytic reactions.1,2) Thus, OH radicals are used as an indicator for catalytic activity.
Gold is an inert material, but nanoscale gold particles exhibit high catalytic activity. This high catalytic activity originates from a high density of dangling bonds on surface steps and kinks and from surface plasmon resonance (SPR) with nano-ordered metal materials. Dangling bonds are uniquely active because they have an unsatisfied valence. Therefore, chemical reactions can be carried out on the surface of the nanostructure, which typically consists of numerous kinks and steps. Moreover, because SPR on nano-ordered metals may enhance the local electric fields, it has been applied in catalytic reactions on nano-ordered metal materials.
Nanoporous gold (NPG) has nano-ordered pores and ligaments. The atomic surface structure of NPG has Au (111) facets on terraces and Au (110) facets on steps.3) Because of this nano-ordered morphology, many studies have reported that NPG also exhibits catalytic activity similar to that of nanoparticles.3–16) NPG is easy to fabricate by using nitric acid to etch away the Ag from Au–Ag alloy films. The pore size and composition of NPG is easy to control by adjusting the etching time and temperature of the nitric acid. Thus, we can obtain NPG with a specific pore size and amount of Ag. This simple and easy preparation process is the one of the attractive features of NPG.
Metal nanoparticles with diameters smaller than 15 nm show the greatest activity in oxidation reactions.6,7) NPG with a pore size of 10–50 nm also shows catalytic activity in similar reactions.8,9) Qian et al. reported that surface-enhanced Raman scattering (SERS) of NPG was observed with nanopore sizes of 5–700 nm, and the strongest SERS enhancement was observed with 5–10 nm nanopores.9) These results suggested that the resonance properties are strongly dependent on the size of the nano-ordered metal materials. Therefore, the catalytic activity of such nano-ordered metal materials depends on the particle size.
In addition to size, the composition of nano-ordered metal materials is also important for catalytic activity. In previous works, it was found that the oxidation of CO by NPG could be considerably increased if the Ag content in the NPG was increased from 0.6 to 11.4 at%.3,7,10,11) A similar effect was reported with Au nanoparticles containing Ag, where the Ag content also enhanced the oxidation of CO.13,14) That is, the catalytic activity was proportional to the Ag content in the samples. Furthermore, it was reported that molecular oxygen (O2) was able to physisorb onto Au, but dissociation was not observed, whereas both adsorption and dissociation of O2 were observed on Ag.15) The reason for this is that the O2 desorption sites on Ag are responsible for raising the desorption enthalpy to be higher than that on Au surfaces. Specifically, the desorption enthalpy of O2 was 11 kJ/mol on the Au (111) surface and 20–30 kJ/mol on the Ag (111) and Ag (110) surfaces.10) This means that once O2 adsorbs onto a Ag site, it will remain there until dissociation. Moreover, Doremus reported that Ag nanoparticles exhibit SPR.16) Tatsuma et al. reported that the dispersion of Ag nanoparticles onto a TiO2 surface enhanced the photocatalytic activity of TiO2.17) These results indicated that the composition of the surface is important for selective catalytic reactions.
One problem with TiO2 is the decrease in photocatalytic activity caused by surface recombination of electron and hole pairs. Surface recombination reduces the production efficiency of OH radicals. Thus, it is necessary to solve the problem of the surface recombination of electron and hole pairs. Several studies have reported that an appropriate nano-ordered co-catalyst is able to effectively reduce the surface recombination of electron and hole pairs, and is also able to significantly enhance the photocatalytic activity by SPR.18,19)
Furthermore, SPR of nano-ordered metal materials has been shown to promote photocatalytic reactions of TiO2 carried out under visible and UV light. For instance, it has been reported that SPR on nanoparticles and nanorods enhances the photocatalytic efficiency of TiO2.6,17,19–23) In addition, after Au nanoparticles are photoexcited as a result of SPR, charge separation is accomplished by the transfer of photoexcited electrons from the Au nanoparticles to the TiO2 conduction band; simultaneously, compensative electrons from a donor in the solution are transferred to the Au nanoparticles. In contrast, in previous studies of the effect of Ag in NPG, the oxidation of CO was enhanced with NPG containing a large amount of Ag, in which case residual Ag on the NPG surface provides desorption sites for O2.11) From this result, we expect that a nanocomposite material composed of NPG and TiO2 (TiO2/NPG) with a large amount of Ag will have a similar effect on OH radical production like that observed in previous studies of Au nano-particles.
For the reasons mentioned above, the photocatalytic activity of nano-ordered materials is strongly dependent on their surface morphology, size, and composition. As such, we expect that the coexistence of TiO2 particles and NPG films will enhance the production of OH radicals. Moreover, the pore size of the NPG will determine the catalytic activity of TiO2/NPG, similar to the enhancement seen with metal nanoparticles on TiO2. The composition of NPG may also exert an influence on the photocatalytic activity of TiO2.
In this study, we present a simple technique for the preparation of a composite material composed of NPG and TiO2 nanoparticles. Moreover, the results of this study illustrate that, in conjunction with TiO2, NPG may act as a promoter to produce OH radicals. Furthermore, the results suggest that this effect depends on the pore size of the NPG, as well as the amount of Ag.
The average pore diameter and amount of residual Ag of the NPG was controlled by adjusting the dealloying temperature and time.24) NPG films were fabricated by chemically dealloying an approximately 100-nm-thick Au35Ag65 (at%) leaf (IMAIKINPAKU Co., Kanazawa, Japan) in a solution of 70 mass% HNO3 (Wako Pure Chemical Industries, Osaka, Japan) for 2–300 min at 0 and 40°C. The as-prepared samples were then carefully rinsed 3 times with distilled water (18.2 MΩ-cm) for 15–20 min (per rinse) to remove residual nitric acid. Finally, the NPG films floating on the distilled water surface were scooped off with glass plates, and fixed onto glass plates by physical adsorption.
The ligament/pore structure was examined with a scanning electron microscope (SEM; JIB-4600F, JEOL, Japan). The average pore diameter was controlled to be between 10 and 80 nm (Fig. 1(a)–(e)). The average sizes of the nanopores were measured by using a rotationally averaged fast Fourier transform (FFT) method.25) The FFT method analyzes the periods of the black (pores) and white (ligaments) areas in SEM images containing 200–300 ligaments/pores. Briefly, the SEM images were transformed to FFT power spectra that contained scattering peaks, which corresponded to the characteristic length scales of the bicontinuous nanostructure. The pore size was defined by the equivalent diameters of the nanopores or gold ligaments and was estimated as half of the characteristic length scale.
Plane-view SEM images of the NPG surface with pore sizes of (a) 10 nm, (b) 20 nm, (c) 30 nm, (d) 40 nm, and (e) 50 nm, and (f) NPG with TiO2.
The composition of the NPG film was determined by using energy dispersive spectroscopy (EDX; Standard, EDAX Inc., Japan) with an attached scanning electron microscope (S-3400N, Hitachi, Japan), which was operated at 15 kV. The composition of Ag was controlled by the dealloying time and temperature of the HNO3. The average amount of Ag in the NPG films was 0–40 at%. The distribution of residual Ag was uniform in the NPG.12)
Commercially available TiO2 (NIPPON AEROSIL Co., Degussa P25, 1.0 ± 0.05 mg) particles with a 4:1 ratio of anatase and rutile species were used as the photocatalytic material. The spin trapping agent utilized in this study was 5-(2,2-dimethyl-1,3-propoxy cyclophosphoryl)-5-methyl-1-pyrroline N-oxide (CYPMPO; 5 × 10−3 mol/L, Radical Research Inc.). TiO2 (1 mg) and CYPMPO (10 mM) were mixed in pure water at 25°C to form the standard solution. Then, we assembled the NPG and TiO2 composite material (TiO2/NPG) by immersing the NPG film in the standard solution (Fig. 1(f)). The X-ray diffraction (XRD) pattern in Fig. 2 shows obvious anatase and rutile peaks, indicating the crystallization of two kinds of TiO2.
XRD profile of TiO2 particles (Degussa P25).
After immersing the glass plate fixed NPG film in the standard solution, the mixture of NPG and the standard solution was irradiated with black light (365 nm; FL10 BLB, TOSHIBA) for 5 min at 25°C. The OH radicals were trapped with CYPMPO as stable spin trap adducts (CYPMPO-OH) at 25°C. X-band electron spin resonance (ESR) spectroscopy (TZ-300, JEOL) was used to estimate the amount of OH radicals generated. The CYPMPO-OH concentration was determined by comparing the obtained signal intensity with that of a standard solution. 100 µL of the solution (without NPG) from the mixture of NPG and the standard solution was collected using a micro-glass tube (Drummond Scientific Co., Microcaps, length 116 mm, outer diameter 1.44 mm, inner diameter 1.05 mm), and measurement by ESR was started 2 to 5 min after collection. The ESR spectrometer settings were as follows: frequency, 9.4 GHz; microwave power, 10.0 mW; field modulation width, 0.063 mT; field scan width/rate, 335.5 ± 7.5 mT/4 min; time constant, 0.1 s; temperature, 293 K.
Figure 3 shows a typical ESP spectrum obtained 5 min after irradiation with black light at 25°C. The peaks labeled as #1 and #2 are derived from OH radicals and we calculated the amount of OH radicals generated from the heights of these peaks.
Typical ESR spectrum of CYPMPO-OH adduct generated during irradiation of black light.
At first, we obtained OH radicals from an Au–Ag alloy film and an NPG film to confirm their ability to produce OH radicals without TiO2 particles. Because NPG has Au (111) facets on its terraces, there is the possibility that water could be activated to transiently generate OH species on the Au (111) surface when oxygen is present.4,5) However, we observed that OH radicals were not generated with the Au35Ag65 alloy film or NPG film (less than 5 at% Ag, pore size of 30–40 nm) alone upon irradiation (results not shown). On the other hand, composite TiO2/NPG films exhibited enhanced production of OH radicals.
Figure 4 shows the OH radical generation in the presence of TiO2/NPG and illustrates the effect of the pore size of NPG when the Ag composition was less than 5 at%. To normalize the data, the signal intensity was divided by the amount of NPG and the intensity of the standard solution. For comparison with Fig. 1, the mean value and the error of the diameter within the range of diameter ±5 nm are shown in Fig. 4. Moreover, the mean value and the error range of the reaction intensity within the range of the diameter ±5 nm are shown in Fig. 4 simultaneously with the diameter size. The amount of OH radicals generated in the presence of TiO2/NPG for nearly all pore sizes was greater than that obtained in the presence of TiO2 alone (the dashed straight line in Fig. 4 represents the signal intensity of TiO2 alone.). Furthermore, NPG with pore diameters of 20–40 nm exhibited high catalytic activity.
Effect of the pore size of NPG with less than 5 at% of Ag on OH radical generation.
Figure 5 shows the effect of the amount of residual Ag in TiO2/NPG on OH radical generation. The amount of OH radicals that were generated was proportional to the amount of residual Ag in the NPG film. The data of the low Ag composition ratio (less than 5 at%) of Fig. 5 re-records the average of all the data shown in Fig. 4 and its error.
Effect of the amount of residual Ag in NPG on OH radical generation.
The results in Fig. 4 illustrate the effect of the composite material with TiO2 and NPG on the generation of OH radicals. That is, these results suggest the promoting effect of NPG on the production of OH radicals with TiO2. Similar enhancement with Au nanoparticles and nanorods has been reported previously.26) One of the possible processes of OH radical production involves first the generation of H2O2 molecules by two-electron reduction of O2 and then the subsequent production of OH radicals from H2O2 by reaction with electrons or holes, because noble metals such as Au serve as an electron pool and also as a catalyst for O2 reduction. (Process A in Fig. 6)
Model of the process of OH radicals generation on the surface of TiO2/NPG.
Furthermore, the peak of catalytic activity occurred at pore diameters of 20–40 nm. Similar size dependence of NPG was reported previously, when an enhancement was observed with materials with a pore size of around 20 nm. Lang et al. reported the pore size dependence of the plasmon peak of NPG in water within the range of 10–50 nm.27) These results suggested a promoter effect and size dependence of NPG for the photocatalytic reaction of TiO2 by SPR.
Moreover, the high catalytic activity of materials with pore sizes of 20–40 nm observed in this work suggests that NPG effectively maintains a high dispersion of TiO2 particles and adhesion between NPG and TiO2 due to its porous structure. Ohno et al. reported that the size of diameter for TiO2 is about 25 nm.28) In the company catalog, the primary particle size is reported to be about 21 nm. According to these informations, the minimum particle size of TiO2 is 20 nm. The reason for the high reaction intensity at 20 to 40 nm is considered to be the result of the TiO2 particles penetrating into the pores of NPG and adhering to the pore surface and preventing the aggregation of the TiO2 particles. In other words, it is considered that the photocatalytic activity of TiO2 was significantly enhanced as a result of the prevention of the reduction of the reaction area and the reduction of the surface recombination of electron and hole pairs.
A cause of the magnitude of the error is the difference in the amount of contact between TiO2 and NPG in each sample. Since the NPG film is simply immersed in the standard solution, whole amount of TiO2 is not fixed to the NPG film. However, since the area of NPG is made constant by normalization, the amount of TiO2 fixed on NPG is considered to be a certain amount to some extent.
The reason of the error is larger as the average pore size is larger is considered as follows. The size distribution width of pore size increases with the growth of pore (experimental results are not shown here). Since the reaction intensity is dependent on the pore size, it is considered that the intensity distribution becomes larger as the size distribution width becomes larger.
Although the detailed mechanism remains to be elucidated, the influence of Ag must be discussed. The results in Fig. 5 suggest an enhancement of OH radical generation owing to Ag on the surface of NPG. The first thing to be considered is that oxygen atoms are generated on Ag sites by the dissociative adsorption of O2, because Ag on the NPG surface is able to enhance oxidative catalytic activity.4,11) Next, the surface of NPG has Au (111) facets on its terraces.3) Moreover, Quiller et al. reported the formation of transient OH radicals from the reaction of water with atomic oxygen on Au (111) surfaces.4) Considering these reports along with our results, it is suggested that Ag on the NPG surface enhances the generation of OH radicals, because oxygen atoms are generated on Ag sites by the dissociative adsorption of O2, and atomic oxygen can generate OH radicals on the Au (111) surfaces of NPG. (Process B in Fig. 6)
Furthermore, we suppose that the high electrical conductivity of Ag effectively reduces the surface recombination of electron and hole pairs for TiO2.
Ag clusters also have the possibility of inducing SPR effects with TiO2 based on previous results.23) However, we suppose that the effect of SPR from Ag clusters is small in this work, because of the size of the Ag clusters in this work is unknown and they might be irregular.
Moreover, Ueno et al. reported water oxidation by composite materials consisting of Au nanorods and TiO2 under irradiation at visible to near-infrared wavelengths.5) Their work suggested that TiO2/NPG has the ability to catalyze water oxidation from visible to near-infrared wavelengths. From these results, we suppose that suitably sized Ag clusters on TiO2/NPG can effectively produce OH radicals under irradiation from visible to near-infrared wavelengths.
As for the synergy between amount of Ag and pore size, the data was not enough to define the conclusion. However, the same dependence on the size as in the sample having a small amount of Ag was observed. Furthermore, even if the pore size was the same, NPG with the large amount of Ag has the higher productivity of OH radicals, and it had tendency to have the more size dependence.
In summary, we were able to assemble NPG and TiO2 composite materials by a simple technique based on immersing of an NPG film in a standard solution. We observed that the generation of OH radicals depends on the pore size of the NPG and the amount of residual Ag. NPG films with pore diameters of 20–40 nm exhibited high catalytic activity, owing to SPR and conduction between the NPG and TiO2 particles with an average diameter of 20 nm. When the NPG had a high amount of Ag, the amount of OH radicals produced was proportional to the amount of Ag. These results suggested that the generation of OH radicals by TiO2/NPG can be tuned by modifying the pore size and the amount of Ag. Therefore, TiO2/NPG may serve as a valuable photocatalyst and NPG can function as a mold and co-catalyst for TiO2 nanoparticles.
This work was performed under the Inter-university Cooperative Research Program of the Institute for Materials Research, Tohoku University (Proposal Nos. 11K0072, 12K0100, and 13G0042). We also acknowledge the technical support from the Hi-tech Research Center of Tohoku Gakuin University and Professor Teruyoshi Awano.
