Precious metal-like oxide-free copper nanoparticles: high oxidation resistance and geometric structure

Abstract

Cu nanoparticles have attracted much attention as an alternative to precious metal Au- and Pt-based materials. However, they have a problem with stability and get oxidized easily. Moreover, their synthesis is carried out with hazardous chemicals and in conditions, which is not environmentally friendly. Thus, both improvement in oxidation resistance and establishment of a novel synthesis method under mild conditions are desired for further applications. To solve these problems, we synthesized Cu nanoparticles from copper(II) acetate by photoreduction at room temperature in the presence of kaolinite, a layered clay mineral with high gas barrier properties. We obtained disk-like metallic Cu nanoparticles adsorbed on kaolinite without oxides. The nanoparticles showed high oxidation resistance in the solid phase. The powder sample was stable and did not get oxidized at all even after 6 months on exposure to fresh air in spite of no apparent protectants, which was revealed by infrared spectroscopy. Namely, the obtained Cu nanoparticles showed a precious metal-like property. On the other hand, the as-prepared Cu nanoparticles still showed the reactivity with fresh air in the suspension. This indicates that the adsorption on kaolinite did not spoil the reactivity of the Cu nanoparticles. Therefore, the present study revealed that the combination of photoreduction and addition of a layered clay mineral would be a promising environmentally friendly way to synthesize oxidation–resistant metal nanoparticles.

---

Introduction

Metal nanoparticles have been investigated due to their high catalytic activity. In particular, considerable efforts have been devoted to synthesize Pt and Au nanoparticles.^1–5 However, the cost of such precious metals is so high that Cu nanoparticles have recently attracted much attention as cheaper alternatives.^6–14 One of the important differences between noble and alternative non-noble metal nanoparticles is the oxidation resistance; Cu is easily oxidized to Cu₂O or CuO.^7,9,10,15,16 To avoid oxidation, numerous methods have been proposed. As a result, it has recently been reported that silica-coating method successfully solved this problem, even though heat-treatment has still been required to protect the Cu nanoparticles firmly.⁹ Moreover, the synthesis of Cu nanoparticles has another problem in that it requires the usage of high reaction temperature and hazardous chemicals as a reductant and a solvent, which are not environmentally friendly. Since Cu is abundant in nature and so-called green syntheses are becoming increasingly important,^17,18 establishment of a synthesis method under mild conditions is significant from the viewpoint of green chemistry.

Previously, our research group has developed a method to synthesize Cu nanoparticles by photoreduction from copper acetate solution in the presence of polyvinylpyrrolidone (PVP).⁶ On exposure to fresh air, a back reaction from the Cu nanoparticles to copper acetate has occurred quickly, followed by regeneration of the Cu nanoparticles by further photoirradiation. For applications of such regenerative Cu nanoparticles, the improvement in oxidation resistance is indispensable. In this regard, the usage of an additive with high gas barrier properties seems a promising way. Additionally, if the Cu nanoparticles adsorb on such an additive without a capping reagent, it is expected to improve the physico-chemical properties of the nanoparticles, including their catalytic activity. Such a method is in contrast to the silica-coating method, which may decrease the catalytic activity of the metal nanoparticles.

For these reasons, we have focused on kaolinite (KL), a layered clay mineral as an additive, because we expect the flat clay surface to be a good adsorbent of Cu nanoparticles. In addition, layered clay minerals are known to have high gas barrier properties,^19–21 which not only stabilizes the Cu nanoparticles but also may make it possible to determine their production mechanism by photoirradiation.

In the present study, we have synthesized Cu nanoparticles on KL, and have tested the oxidation resistance on exposure to fresh air to evaluate advantage of layered clay minerals. We have also revealed the production mechanism both by identifying the reaction intermediate and by analyzing the reactant kinetically.

Experimental

Copper nanoparticles were synthesized by a photoreduction method, similar to that reported previously.⁶ In brief, copper(II) acetate monohydrate (purity > 99.0%) was purchased from Wako Pure Chemical Industries, and an 8 mmol L⁻¹ aqueous solution of copper acetate containing 10% ethanol was prepared. Ten milligrams of KL (JCSS-1101 reference clay sample of the Clay Science Society of Japan, Kanpaku mine, Tochigi Prefecture, Japan) was added to 10 mL of the mixed solution. The obtained suspension was irradiated by UV light coming from a super high-pressure mercury lamp (Ushio, BA-H502) under vigorous stirring.

For scanning electron microscopy (SEM) measurement, the suspension was rapidly dropped on a collodion-coated copper grid. SEM images of the dried grid were observed with a field-emission SEM (Hitachi High-Technologies, S-5500) operated at 5 kV.

UV-vis-NIR absorption spectra were measured with an UV-vis-NIR spectrometer (Shimadzu, UV-3600) using a rectangular quartz cuvette having an optical length of 10 mm.

The as-prepared photoirradiated suspension was dropped on a glass plate under nitrogen atmosphere and dried in vacuum. X-ray diffraction (XRD) measurement of the obtained powder samples was performed with an X-ray diffractometer (Rigaku, Smartlab) with CuKα radiation (0.154 nm). The oxidation state of the powder sample was further analyzed by X-ray photoelectron spectroscopy (XPS) with an X-ray photoelectron spectrometer (AXIS-His, Kratos Analytical) using monochromatic X-ray AlKα radiation (1486.6 eV). The as-prepared photoirradiated suspension was also analysed by Cu K-edge XAFS (JASRI, Hyogo, Japan). The Cu K-edge XAFS spectra were analysed by the program package REX2000 (Rigaku Co.). Model parameters for curve-fitting analysis (back scattering amplitude and phase shift) was extended from an EXAFS oscillations observed for Cu foil.

Powder samples were also diluted with KBr, and infrared (IR) spectra were measured with an FTIR spectrometer (JASCO, FT/IR-4100). The spectral resolution was 1 cm⁻¹, and the number of accumulations was 32.

The photoirradiated suspension was exposed to fresh air under vigorous stirring, followed by a further photoirradiation test to investigate the oxidation resistance and the regenerative ability.

Results and discussion

Geometric structure of Cu nanoparticles

When the copper acetate suspension was irradiated by UV light, successive color changes were observed, as shown in Fig. 1: light blue, yellow, orange, brown, gray, and deep red. The initial light blue and the final deep red colors are due to copper acetate and surface plasmon resonance (SPR) of the Cu nanoparticles, respectively. However, the latter was much deeper than that observed for spherical Cu nanoparticles in our previous study,⁶ suggesting that the obtained Cu nanoparticles were not spherical but anisotropic.

Fig. 1 Photographs of the photoirradiated suspension. Number indicates photoirradiation time in hours.

Fig. S1† shows a SEM image of the suspension photoirradiated for 24 h. Small disk-like particles are adsorbed on a large flat plate. Since KL has a flat surface with a μm-sized diameter, it was concluded that disk-like Cu nanoparticles (Cu dNPs) were produced and adsorbed on KL. The diameter of the Cu dNPs was about 80 nm. All the Cu dNPs were adsorbed on KL, indicating that its flatness proved adequate for the KL surface to act as an adsorbent.

Fig. S2† shows a UV-vis-NIR extinction spectrum of the nanocomposite of Cu dNPs and KL (Cu dNP–KL, hereafter) suspension. To avoid the decrease in sensitivity due to scattering, the suspension was diluted by 5 times. A broad SPR band appeared at 596 nm, which was consistent with the peak wavelength of Cu nanodisks on a glass substrate as reported previously.¹⁵ Compared to spherical Cu nanoparticles appearing at 570 nm sharply,^6,9,22 the SPR band of Cu dNPs in the present study was red-shifted and relatively broad, similar to other metal nanodisks and nanoplates reported previously.^23–26 Therefore, we concluded that Cu dNP–KL was produced by UV irradiation.

We assumed that the addition of KL had affected the formation of Cu dNPs. To figure out the function of KL, the copper acetate solution without KL was irradiated by UV light. As a result, large copper aggregates with sub-mm size were obtained, as reported previously.²⁷ This result confirms that KL is the adsorbent of Cu dNPs and prevents them from aggregating.

The photoirradiated suspension separated into two phases when the stirring was stopped. The color of the precipitate was deep red, while the supernatant was totally transparent. These findings suggest that Cu²⁺ was reduced by UV irradiation, and the obtained Cu dNPs were adsorbed on KL. UV-vis-NIR absorption spectra of the supernatant before and after UV irradiation for 24 h were measured, as shown in Fig. S3.† The band at 780 nm originated from Cu²⁺ decreased by UV irradiation, and almost completely disappeared at 24 h. Thus, it was confirmed that Cu²⁺ was reduced by UV irradiation, resulting in the formation of the Cu dNPs. In contrast, when the sample was stirred for 24 h in darkness, the obtained spectrum was almost consistent with that before UV irradiation, as shown in Fig. S3.† This result indicates that Cu²⁺ was not adsorbed on KL during the UV irradiation.

Oxidation state of Cu nanoparticles

Oxidation state of the photoirradiated sample was examined by XRD, as shown in Fig. 2. According to JCPDS file no. 4-0836, the three strong peaks at 43.3°, 50.4° and 74.1° correspond to the Miller planes (111), (200) and (220) of the crystalline Cu metal, respectively. All other peaks were assignable to KL, and no peaks of impurities were observed. In Fig. 2(a), no peaks were observed at 42.3° and 48.7° originating from Cu₂O and CuO, respectively (JCPDS files no. 5-0667 and 48-1548). The fact that only metallic Cu dNPs were obtained without oxides by UV irradiation for 24 h is notable since the surface of Cu nanoparticles is known to be oxidized easily, resulting in a core–shell structure composed of Cu-core and Cu₂O-shell.^7,28,29 Also, in contrast to previous studies, where Cu₂O nanoparticles were produced by the reduction of copper acetate,^30,31 Cu²⁺ was reduced to Cu⁰ completely in the present study. Furthermore, the observed intensity ratio of the strongest peak belonging to the (111) plane to the peak belonging to the (200) plane was significantly higher than that for spherical Cu nanoparticles reported in the previous studies.^8,9,16,32 This high intensity ratio agrees with the disk-like structure because the top crystal plane of the Cu nanodisks has a predominant (111) orientation,³³ and a similar high intensity ratio was also reported in Au and Ag nanoplates.^34–37

Fig. 2 X-ray diffraction (XRD) patterns of (a) Cu dNP–KL, (b) and (c) Cu dNP–KL exposed to fresh air for 1 and 6 months, respectively.

Oxidation resistance of Cu nanoparticles

Oxidation resistance of Cu dNPs was evaluated in the solid phase. The powder sample was exposed to fresh air in darkness at room temperature to evaluate its oxidation resistance. As a result, the XRD measurement revealed that the sample after 1 month did not show any peaks of Cu₂O or CuO, as shown in Fig. 2(b).

Absence of Cu₂O and CuO was also confirmed by XPS. Fig. 3 shows X-ray photoelectron spectra of Cu 2p of Cu dNP–KL. Two peaks were observed at 932.2 and 952.0 eV. The former peak is ascribed to metallic Cu 2p_3/2 and the energy difference between the two peaks corresponds to a spin–orbit separation between Cu 2p_3/2 and Cu 2p_1/2. Both spectra of the samples right after the preparation and after the air exposure for 1 month were well fitted by one Gaussian–Lorentzian curve, indicating that even amorphous Cu₂O was not produced by the air exposure. This result was further confirmed by XAFS measurement. Fig. 4 and 5 show XANES and Fourier-transformed EXAFS spectra of Cu dNP–KL and Cu foil. In Fig. 4, since absorption edges and peak profiles in the XANES spectra of Cu dNP–KL both right after the preparation and after the air exposure for 2 months were almost consistent with those of the Cu foil, it was found that the oxidation state of Cu dNP–KL were metallic. In the EXAFS spectra shown in Fig. 5, since the strongest peaks of Cu dNP–KL both right after the preparation and after the air exposure for 2 months were distributed in the same range as that of the Cu foil, the peaks can be assigned to the scattering from the first nearest-neighbored Cu atoms. All these findings indicate that Cu dNP–KL has extremely high oxidation resistance in the solid phase. In fact, no peaks of Cu₂O were observed in the XRD pattern even after the air exposure for 6 months as shown in Fig. 2(c). Additionally, the intensity ratio of the strongest peak belonging to the (111) plane to the peak belonging to the (200) plane did not change, suggesting that the anisotropy was still maintained.

Fig. 3 X-ray photoelectron spectra of Cu 2p of (a) Cu dNP–KL and (b) Cu dNP–KL exposed to fresh air for 1 month.

Fig. 4 XANES spectra of (a) Cu dNP–KL, (b) Cu dNP–KL exposed to fresh air for 2 months, and (c) Cu foil.

Fig. 5 Fourier-transformed EXAFS spectra of (a) Cu dNP–KL, (b) Cu dNP–KL exposed to fresh air for 2 months, and (c) Cu foil.

One may think that the high oxidation resistance of the Cu dNPs resulted from the acetate ion (OAc⁻) protecting the Cu dNPs firmly, which is similar to a previous study where the Cu nanoparticles were protected by oleic acid.⁸ To examine the presence of the OAc⁻, IR spectra of KL and Cu dNP–KL were measured, as shown in Fig. S4.† Since no C=O stretching band of AcO⁻ was observed in the region between 1600 and 1800 cm⁻¹, it was concluded that the Cu dNPs were not protected by AcO⁻. All IR bands were assignable to KL except for a weak band around 640 cm⁻¹ assigned to alunite contained in KL as an impurity, whose relative intensity is quite weak, indicating that the amount of the impurity should be almost negligible. A sharp band at 668 cm⁻¹ was due to CO₂. Thus, the obtained Cu dNPs were found to be adsorbed on KL without a capping reagent. In other words, the high oxidation resistance of Cu dNPs is due to the presence of KL with its high gas barrier properties.

Reactivity of Cu nanoparticles

The reactivity of Cu dNP–KL in the suspension was also evaluated. The suspension was exposed to fresh air by stirring vigorously in an open system to avoid precipitation. As a result, successive color changes were observed, shown in Fig. S5.† The deep red color gradually faded away within 60 min. At 8 h of the air exposure, a pale light blue color was observed in the supernatant, when the stirring was stopped. Thus, it was found that the back reaction from Cu dNPs to copper acetate occurred, which was confirmed by UV-vis-NIR spectra shown in Fig. S6.† Therefore, we concluded that the addition of KL did not prevent Cu dNPs from being oxidized to Cu²⁺.

Production mechanism of Cu nanoparticles

As shown in Fig. 6, Cu₂O was detected as a reaction intermediate. This indicates that the production of Cu dNPs proceeds in the following two-step reaction denoted as

Fig. 6 Time-dependent changes of XRD patterns of Cu dNP–KL as a function of UV irradiation time.

In the following sections, these reaction steps are elucidated quantitatively.

Photoreduction of copper acetate as the first reaction step

The time evolution of the reactant, Cu²⁺ is plotted using solid circles in Fig. 7. The abundance of Cu²⁺ was estimated from the UV-vis-NIR absorption band at 780 nm. Since the obtained value seemed to decrease exponentially, it was assumed that Cu²⁺ was reduced by a first-order reaction upon UV irradiation. The corresponding rate equation can be derived from the general kinetic theory as follows:where the concentration of Cu²⁺ in the suspension and the rate constant are denoted as [Cu²⁺] and k, respectively. By solving the differential eqn (1), the eqn (2) is obtained as follows:where [Cu²⁺]₀ is the initial concentration of copper acetate. The value of k was determined to be (3.3 ± 0.2) × 10⁻⁵ s⁻¹ by applying the least-square fitting process. The half-life period was calculated to be 2.1 × 10⁴ s. Since the experimental data points lie closed to the theoretical curve shown in Fig. 7, we concluded that Cu²⁺ was reduced to yield Cu₂O by a first-order reaction upon UV irradiation. It should be noted that the production of Cu₂O results from a photoinduced electron transfer from AcO⁻ to Cu²⁺, according to the previous studies on the photoredox properties of copper(II)–carboxylate complexes.^38–41

Fig. 7 Absorbance and intensity changes at a function of UV irradiation time. The dotted line represents the calculated value of absorbance using the eqn (2).

As described already, the production of Cu₂O as the reaction intermediate was detected in the XRD patterns shown in Fig. 6. One strong and one weak peaks at 36.4° and 42.3°, respectively appeared at 4 h of UV irradiation, corresponding to the (111) and (200) planes of Cu₂O, respectively. Thus, it was suggested that the reduction from Cu²⁺ to Cu⁰ was a stepwise reaction, which proceeded through a reaction intermediate, Cu₂O. In addition, it is reported that the (111) plane is the top crystal plane of Cu₂O nanoplates.⁴² The intensity ratio of the peak belonging to the (111) plane to the peak belonging to the (200) plane was found to be higher in the present study than that of spherical Cu₂O nanoparticles reported previously.^9,30,43,44 Thus, this similarity to the case of Cu dNPs suggests that Cu₂O dNPs were produced by UV irradiation. Furthermore, when the stirring had stopped, the obtained precipitate was colored and the supernatant was pale light blue, indicating that Cu₂O dNPs were adsorbed on KL. Actually, such an adsorption and the production of Cu dNPs also support the production of Cu₂O dNPs, since Cu₂O was immobilized on KL by the adsorption. Therefore, we concluded that the reduction from Cu²⁺ to Cu₂O occurred in solution, followed by the immediate adsorption onto KL to produce Cu₂O dNPs, as shown in Scheme S1(a).†

Formation of Cu nanoparticles as the second reaction step

As shown in Fig. 6, no peaks of Cu⁰ appeared at 43.3° at 4 h of UV irradiation. From 4 to 12 h, the peak intensities of Cu₂O increased, while a very weak peak of Cu⁰ appeared at 12 h. From 12 h to 20 h, the peak intensities of Cu⁰ increased, while those of Cu₂O were almost constant. From 20 h to 24 h, the peak intensities of Cu₂O decreased, and those of Cu⁰ increased correspondingly. These results are consistent with the observations that a gray color was observed from 8 h to 16 h, which turned to deep red from 16 h to 24 h. As for the peaks of Cu⁰, the intensity ratio of the strongest peak belonging to the (111) plane to the peak belonging to the (200) plane was almost constant from 20 h to 24 h. These findings indicate that the anisotropy of the Cu dNPs did not change during the production of the Cu₂O dNPs, which produced Cu dNPs at the final stage of UV irradiation (Scheme S1(b)†).

Fig. 6 shows the time evolution of Cu₂O and Cu⁰ obtained from the XRD peaks at 36.4° and 43.3°, respectively. While the data points of Cu₂O increased in correspondence to the decrease of Cu²⁺, those of Cu⁰ started increasing from 16 h, when sufficient amount of Cu₂O was present in the suspension. Such a long induction period of Cu⁰ clearly supports that Cu dNPs were produced not from Cu²⁺ directly but from the reaction intermediate, Cu₂O dNPs. Therefore, we concluded that Cu²⁺ was reduced to yield Cu₂O dNPs, followed by further photoreduction to Cu dNPs.

Solvent dependency on the production of Cu dNPs was also investigated to confirm the production mechanism. Two experiments were performed, where pure ethanol and water solvent were used. On the one hand, the color changes were much slower in pure ethanol solution than in the water–ethanol mixed solution. This result suggests that the presence of sufficient amount of water is important to produce Cu₂O from Cu²⁺ effectively. On the other hand, the color changes were also slower in pure aqueous solution than those in the mixed solution, and the color at 24 h was gray. The XRD measurement of this gray sample revealed the production of Cu₂O. We assume that ethanol reacts as a sacrificial reagent in the present study, and AcO⁻ may decompose during the photoredox reaction. Indeed, when the sample containing 4 mmol L⁻¹ of copper acetate solution containing ethanol was used to suppress the saturated absorption of AcO⁻ below 350 nm, a weak band was observed around 280 nm. This band was well consistent with that of dilute acetaldehyde aqueous solution, as shown in Fig. S7.† Thus, it was suggested that ethanol is the sacrificial reagent during the photoredox reaction of Cu²⁺. One may think that acetaldehyde also reacts as a sacrificial reagent during the UV irradiation to produce acetic acid. However, considering the fact that the amount of ethanol (17.1 mmol) is much larger than that of acetaldehyde (0.04 mmol at a maximum), it is plausible that ethanol reacts preferentially as the sacrificial reagent. The presence of acetaldehyde was also supported by gas chromatography with a retention time of 44 s. Combined with the obtained results hitherto, the most probable production mechanisms of Cu dNP–KL were proposed in Scheme S1.†

Conclusions

In the present study, Cu nanoparticles were synthesized from copper acetate by photoreduction in the presence of kaolinite, a layered clay mineral. The obtained Cu nanoparticles were metallic, disk-shaped and adsorbed on kaolinite. The powder sample of the disk-like Cu nanoparticles showed extremely high oxidation resistance due to the high gas properties of kaolinite. In other words, the obtained Cu nanoparticles behaved as precious metal nanomaterials. It is notable that the reactivity of the Cu nanoparticles suspension with fresh air was not spoiled by the adsorption on kaolinite. The mechanism of the formation of Cu nanoparticles was also elucidated by analyzing UV-vis-NIR spectra and XRD patterns kinetically. It was found that Cu²⁺ was reduced by a first-order reaction to produce Cu₂O as an intermediate, which was assumed to be disk-shaped and reduced to the disk-like Cu nanoparticles by further UV irradiation.

Acknowledgements

The XAFS measurements were performed at the BL01B1 of Spring-8 with the approval of the Japan Synchrotron Radiation Research Institute (JASRI) (Proposal No. 2015B1068). This work was supported by MEXT KAKENHI Grant Number 25248004 and by JSPS KAKENHI Grant Number 25410028.

Notes and references

1. Z. U. Rahman, T. Zhang, S. Cui and D. Wang, RSC Adv., 2015, 5, 99697.
2. M. Wang, Z. Zhang and J. He, Langmuir, 2015, 31, 12911.
3. V. Latour, A. Maisonnat, Y. Coppel, V. Collière, P. Fau and B. Chaudret, Chem. Commun., 2010, 46, 2683.
4. C. Tojo, M. Dios, D. Buceta and M. A. López-Quintela, Phys. Chem. Chem. Phys., 2014, 16, 19720.
5. H. Lv, Z. Xi, Z. Chen, S. Guo, Y. Yu, W. Zhu, Q. Li, X. Zhang, M. Pan, G. Lu, S. Mu and S. Sun, J. Am. Chem. Soc., 2015, 137, 5859.
6. N. Nishida, A. Miyashita, N. Hashimoto, H. Murayama and H. Tanaka, Eur. Phys. J. D, 2011, 63, 307.
7. N. Nishida, A. Miyashita, T. Tsukuda and H. Tanaka, Chem. Lett., 2013, 42, 168.
8. S. Jeong, S. H. Lee, Y. Jo, S. S. Lee, Y.-H. Seo, B. W. Ahn, G. Kim, G.-E. Jang, J.-U. Park, B.-H. Ryu and Y. Choi, J. Mater. Chem. C, 2013, 1, 2704.
9. S. Shiomi, M. Kawamori, S. Yagi and E. Matsubara, J. Colloid Interface Sci., 2015, 460, 47.
10. D. Talukdar, G. Das, S. Thakur, N. Karak and A. Thakur, Catal. Commun., 2015, 59, 238.
11. B. J. Borah, S. J. Borah, L. Saikia and D. K. Dutta, Catal. Sci. Technol., 2014, 4, 1047.
12. B. K. Park, S. Jeong, D. Kim, J. Moon, S. Lim and J. S. Kim, J. Colloid Interface Sci., 2007, 311, 417.
13. B. Bagchi, S. Kar, S. K. Dey, S. Bhandary, R. Roy, T. K. Mukhopadhyay, S. Das and P. Nandy, Colloids Surf., B, 2013, 108, 358.
14. R. Reske, H. Mistry, F. Behafarid, B. R. Cuenya and P. Strasser, J. Am. Chem. Soc., 2014, 136, 6978.
15. M. Schwind, S. Hosseinpour, C. Langhammer, I. Zorić, C. Leygraf and B. Kasemo, J. Electrochem. Soc., 2013, 160, C487.
16. G. Das, R. D. Kalita, P. Gogoi, A. K. Buragohain and N. Karak, Appl. Clay Sci., 2014, 90, 18.
17. G. Elango and S. M. Roopan, J. Photochem. Photobiol., B, 2016, 155, 34.
18. K. J. Rao and S. Paria, RSC Adv., 2014, 4, 28645.
19. D. A. Hagen, L. Saucier and J. C. Grunlan, ACS Appl. Mater. Interfaces, 2014, 6, 22914.
20. K. M. Holder, M. A. Priolo, K. E. Secrist, S. M. Greenlee, A. J. Nolte and J. C. Grunlan, J. Phys. Chem. C, 2012, 116, 19851.
21. T. Ebina and F. Mizukami, Adv. Mater., 2007, 19, 2450.
22. J. Bárta, M. Pospíšil and V. Čuba, J. Radioanal. Nucl. Chem., 2010, 286, 611.
23. Y. Ide, M. Matsuoka and M. Ogawa, J. Am. Chem. Soc., 2010, 132, 16762.
24. S. S. Shankar, A. Rai, A. Ahmad and M. Sastry, Chem. Mater., 2005, 17, 566.
25. P. Hanarp, M. Käll and D. S. Sutherland, J. Phys. Chem. B, 2003, 107, 5768.
26. J. E. Millstone, S. Park, K. L. Shuford, L. Qin, G. C. Schatz and C. A. Mirkin, J. Am. Chem. Soc., 2005, 127, 5312.
27. J. Long, J. Dong, X. Wang, Z. Ding, Z. Zhang, L. Wu, Z. Li and X. Fu, J. Colloid Interface Sci., 2009, 333, 791.
28. S. K. Movahed, M. Dabiri and A. Bazgir, Appl. Catal., A, 2014, 481, 79.
29. L. Xu, C. Srinivasakannan, J. Peng, M. Yan, D. Zhang and L. Zhang, Appl. Surf. Sci., 2015, 331, 449.
30. H.-R. Jung, S.-H. Park and W.-J. Lee, Mater. Chem. Phys., 2012, 135, 340.
31. P. Wang, H. Wu, Y. Tang, R. Amal and Y. H. Ng, J. Phys. Chem. C, 2015, 119, 26275.
32. Y. Luo, Y. Tu, Q. Ren, X. Dai, L. Xing and J. Li, J. Solid State Chem., 2009, 182, 182.
33. C. Salzemann, J. Urban, I. Lisiecki and M.-P. Pileni, Adv. Funct. Mater., 2005, 15, 1277.
34. B. Liu, J. Xie, J. Y. Lee, Y. P. Ting and J. P. Chen, J. Phys. Chem. B, 2005, 109, 15256.
35. X. Jiang, Q. Zeng and A. Yu, Nanotechnology, 2006, 17, 4929.
36. Q. Wu, P. Diao, J. Sun, T. Jin, D. Xu and M. Xiang, J. Phys. Chem. C, 2015, 119, 20709.
37. T. Parnklang, B. Lamlua, H. Gatemala, C. Thammacharoen, S. Kuimalee, B. Lohwongwatana and S. Ekgasit, Mater. Chem. Phys., 2015, 153, 127.
38. J. Long, J. Dong, X. Wang, Z. Ding, Z. Zhang, L. Wu, Z. Li and X. Fu, J. Colloid Interface Sci., 2009, 333, 791.
39. J. Y. Morimoto and B. A. Degraff, J. Phys. Chem., 1972, 76, 1387.
40. J. Y. Morimoto and B. A. Degraff, J. Phys. Chem., 1975, 79, 326.
41. P. Natarajan and G. Ferraudi, Inorg. Chem., 1981, 20, 3708.
42. C.-H. Kuo, T.-E. Hua and M. H. Huang, J. Am. Chem. Soc., 2009, 131, 17871.
43. Z. Zhai, X. Guo, Z. Jiao, G. Jin and X.-Y. Guo, Catal. Sci. Technol., 2014, 4, 4196.
44. A. Chakravarty, K. Bhowmik, A. Mukherjee and G. De, Langmuir, 2015, 31, 5210.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra18076k

---