Facile synthesis and the photo-catalytic behavior of core–shell nanorods

Abstract

Sodium niobate nanorods (SNRs) have been synthesized by a facile surfactant free hydrothermal method. To explore their potential for photoelectrochemical water splitting under visible light, core–shell nanorods were fabricated by grafting CdS on sodium niobate nanorods. The TEM analysis shows the formation of sodium niobate nanorods which are in the order of 40 ± 5 nm in width and 1300 ± 100 nm in length. The presence of a thin layer on nanorods, as observed in a TEM image, and XRD and SAD analysis, reveals the grafting of hexagonal CdS on orthorhombic sodium niobate nanorods. This was further confirmed by dual band gap values (E_g: 3.6 for sodium niobate and 2.59 eV for CdS) determined from diffuse reflectance data of the CdS–sodium niobate nanorod sample. The CdS–sodium niobate nanorods show drastic enhancement in the current density (J_an: 7.6 mA cm⁻² at 0.2 V vs. SHE) when irradiated with monochromatic UV light (300 nm), many folds higher than that observed for bare sodium niobate nanorods (J_an: 2.5 mA cm⁻² at 0.2 V vs. SHE), bulk sodium niobate (J_an: 0.6 mA cm⁻² at 0.2 V vs. SHE) and CdS. The conduction band (CB) minima calculations show a downhill offset of the CB edges of CdS–sodium niobate. Such a downhill staggered band gap and smooth lattice matched interface, as shown by HRTEM, seem to facilitate an efficient charge separation followed by a photo-generated e⁻ transfer from the CdS CB to the sodium niobate CB and, therefore, appear responsible for the enhancement of the photocurrent density of CdS–sodium niobate nanorods. This is further corroborated by the time resolved photoluminescence decay measurements which show a longer average decay time (〈τ〉) for CdS–sodium niobate nanorods in the order of 8.06 ns than that for sodium niobate nanorods (6.45 ns). Furthermore, better light harvesting efficiency and incident to photon conversion efficiency (23.91% at 300 nm) observed for CdS–sodium niobate nanorods imply a better photo-generated charge carrier separation than those observed for bare sodium niobate nanorods and bulk sodium niobate. The synthesis of CdS modified sodium niobate nanorods, detailed results on the photoelectrochemical behaviour of CdS modified sodium niobate nanorods and underlying mechanism are presented.

---

Introduction

A variety of catalysts such as semiconductors, organometallic complexes and enzymes are being explored to develop an artificial photosynthesis process/device for splitting water under visible solar radiation. The overall efficacy of such a process/device is limited by the efficiency of the charge separation and rate of electron transfer to the reaction site.^1–3 An efficient charge separation, so that recombination of photo-generated charge carriers is minimized, is still a challenge for the scientific community. An important class of metal oxide semiconductors that have a layered structure has drawn significant attention for their use as photocatalysts.^4–13 The interlayer spacing of these layered metal oxide semiconductors provides reactants (H⁺ and/or H₂O) with accessibility (proximity) to the reaction sites and also allows foreign cations to intercalate so that their properties can further be tuned.^11,12 The incorporation of alkaline metal cations in such layered structures leads to a higher degree of hydration of the interlayer, providing reactants (H₂O or H⁺) access to the reaction sites, or in other words, access to the excitons.^6,11 Various attempts to load NaNbO₃ with Fe, Ni, Co, Ag, and Pt have been made to drive the decomposition of formic acid or the production of H₂.^13–16 However, the use of expensive noble metals and their wide band gap (3.3–3.4 eV – can harvest only UV radiation) curtail their significance for practical applications.

One possible approach to tune the band gap so that NaNbO₃ can harvest visible solar radiation is to attach organic dyes (sensitizers) to it, but these dyes suffer from instability and may undergo oxidation. Another approach to sensitize sodium niobate is to graft narrow band gap semiconducting nanocrystals. In particular, the fabrication of core–shell semiconducting nanostructures has drawn significant attention for photocatalysis applications, as these structures widen the photoresponse/light harvesting region and the interface of the core–shell structure greatly influences the separation of photo-generated charge carriers.^17–21 Thereby, we synthesized CdS modified NaNbO₃ core–shell nanorods to study their potential for photo-electrocatalytic water splitting applications.

In this article a facile synthesis of NaNbO₃ nanorods by a surfactant free hydrothermal method is reported. To explore their potential for photoelectrochemical water splitting under visible light, CdS has been grafted on these nanorods. The performance of CdS modified NaNbO₃ has been examined with respect to NaNbO₃ nanorods, bulk NaNbO₃ and CdS. The detailed results on the structure, photoelectrochemical properties, carrier densities, photo-exciton life times and possible underlying mechanism are presented.

Experimental

Niobium ethoxide (99.95%, Aldrich) was used as a precursor for sodium niobate synthesis. Sodium hydroxide and ethylene glycol were obtained from Merck. To synthesize cadmium sulfide, cadmium acetate and sulfur were obtained from Merck. Dodecylamine (Aldrich) was used as a surface passivation agent. All reagents were used as supplied without any further purification.

Synthesis of NaNbO₃ nanorods

In a typical synthesis, 0.5 g of Nb(OC₂H₅)₅ was mixed with 10 mL of ethylene glycol and subsequently, 1 mL of 20 M NaOH was added dropwise at 40 °C and allowed to stir for 1 h. The reaction mixture was transferred into a Teflon-lined autoclave and allowed to age at 200 °C for 24 h. The product was recovered by centrifugation and washed with distilled water and ethanol. The obtained powder was then subjected to calcination at 550 °C for 4 h. This sample is referred to as SNR in the following text. Bulk sodium niobate was prepared by a conventional solid state method using Nb₂O₅ (Sigma) and Na₂CO₃ (Merck).¹³ Henceforth, this sample is referred to as BSN in the following text.

Synthesis of CdS nanoparticles

To synthesize CdS nanoparticles, a solution containing 4 mmol Cd(CH₃COO)₂·2H₂O, 4 mmol S powder and 10 mL dodecylamine (pre-heated at 50 °C while stirring for 30 min) was transferred into a Teflon-lined autoclave and allowed to age at 220 °C for 10 h. The product was recovered by centrifugation and washed with ethanol. It was then dried at 70 °C overnight.

Grafting of CdS onto NaNbO₃ nanostructures. To grow CdS on sodium niobate nanorods, the solution containing 4 mmol Cd(CH₃COO)₂·2H₂O, 4 mmol S powder and 10 mL dodecylamine (pre-heated at 50 °C while stirring for 30 min) was transferred into a Teflon-lined vessel containing the reaction mixture for sodium niobate as mentioned above. Furthermore, the same reaction steps were followed as mentioned in the sodium niobate nanorod synthesis. This sample is referred to as CdS–SNR in the following text.

Characterization and measurements

The structural characterization (both phase and morphology) of synthesized photo-catalysts was undertaken using a X-ray diffractometer (X'PERT PANalytical) and scanning electron microscope (S-3400N Hitachi). To ensure CdS grafting on sodium niobate and explore microscopic structural details, the samples were subjected to a transmission electron microscopy (TEM) investigation, carried out using a FEI, TECNAI G² F30, S-TWIN microscope operating at 300 kV and equipped with a GATAN Orius CCD camera. High-angle annular dark field scanning transmission electron microscopy (STEM-HAADF) was employed using the same microscope, which was equipped with a scanning unit and a HAADF detector from Fischione (model 3000). The compositional analysis was performed by an energy dispersive X-ray spectroscopy (EDS, EDAX Instruments) attachment on the Tecnai G² F30. Energy-filtered TEM (EFTEM) measurements were carried out using a GIF Quantum SE (model 963). The sample was dispersed in ethanol using an ultrasonic bath, mounted on a carbon coated Cu grid, dried, and used for TEM measurements.

The optical properties (band gap, E_g) were examined using a UV-Vis spectrophotometer (UV-2450 SHIMADZU). The photo-electrochemical behavior of these photo-catalyst samples was studied using a three-electrode potentiostat (Princeton Applied Research). Ag/AgCl and Pt were used as a reference and counter electrode, respectively.

Time-resolved photoluminescence decay measurements were carried out using a time correlated single-photon counting (TCSPC) spectrometer (Edinburgh, OB920). A diode laser (375 nm) was used as the excitation source, and a MCP photomultiplier (Hamamatsu R3809U-50) was used as the detector (response time 40 ps). The instrument response function of the experimental set up is limited by the full width at half maxima (FWHM) of the excitation laser pulse and is 75 ps for the 375 nm source. The lamp profile was recorded by a scatterer (dilute Ludox solution in water) in place of the sample. Time-resolved photoluminescence decay profiles were analyzed by nonlinear least squares iteration procedures using F900 decay analysis software. The quality of the fit is assessed by the chi-square (χ²) values and distribution of residuals.

To undertake the photoelectrochemical studies of the SNRs, CdS–SNRs and BSN, powders of the respective samples were pelletized. These pellets were then converted into electrodes by making ohmic contact with a Cu wire using silver paint and were covered by epoxy. An aqueous solution containing 0.1 M Na₂S and 0.14 M Na₂SO₃ (pH ∼ 12.4) was used as an electrolyte. A monochromator consisting of a 500 W Xenon-mercury lamp was used as a light source for the measurement of current density at different incident wavelengths. A 150 watt Solar simulator (Hi-Tech) equipped with an AM 1.5G filter was used to measure the photocurrent density. Mott–Schottky data were recorded under similar conditions to PEC experiments using an IVIUM electrochemical work station at 1 kHz.

Results and discussion

The XRD pattern recorded for SNRs exhibits the formation of predominantly an orthorhombic phase of sodium niobate (ESI, Fig. S1†). The grafting of CdS on SNRs was confirmed by the appearance of a diffraction peak [101], as observed in addition to the peaks for the orthorhombic phase of sodium niobate, which corresponds to the hexagonal phase of CdS. The BSN sample shows the formation of an orthorhombic phase, (Fig. S1, ESI†). The SEM images recorded for SNRs and CdS–SNRs show the formation of nanorods, Fig. 1(a) and (b). These nanorods have a high aspect ratio and are in the order of 40 ± 5 nm in width and 1300 ± 100 nm in length. The BSN, as expected, shows the formation of big particles with irregular shapes (in the order of 1 μm, Fig. S2, ESI†).

Fig. 1 SEM images of (a) SNRs and (b) CdS–SNRs.

To get structural information of the hybrid CdS–SNR sample, a detailed TEM analysis was performed. A low magnification TEM image shows a thin layer coated on the nanorods (Fig. 2(a)). It implies the grafting of CdS on sodium niobate nanorods. The selected area diffraction (SAD) pattern of a region marked by a dotted circle is shown in Fig. 2(b). The concentric ring consisting of distinct spots is a result of many small crystals and indicates the crystalline nature of heterostructured CdS–SNRs. The measured interplanar spacings from the SAD pattern further confirms the formation of a crystalline orthorhombic phase of sodium niobate and the grafting of hexagonal CdS. The high-resolution TEM image (Fig. 2(c)) and the Fourier filtered image shown in Fig. 2(d) of the region marked by a dotted box in Fig. 2(c) clearly show lattice fringes across the interface between two layers. Indeed the interface appears very smooth and lattice matched. The measured lattice spacing of the CdS layer is 2.44 Å and that of NaNbO₃ is also 2.44 Å. The interplanar spacing of hexagonal CdS (102) is 2.43 Å and that of orthorhombic NaNbO₃ (112) is 2.43 Å. To investigate the chemical composition of the core–shell structure, we performed a STEM-HAADF analysis. It provides a Z-contrast image, where the intensity of scattered electrons is proportional to the square of the atomic number Z. Fig. 2(e) depicts the STEM-HAADF image of a CdS–SNR core–shell nanorod. The spatial distributions of the atomic contents across the CdS–SNR core–shell nanorod were obtained using an energy dispersive X-ray spectroscopy (EDX) line profile. The inset of Fig. 2(e) shows the EDX profiles of Nb, and Cd across line 1.

Fig. 2 Electron microscopy of CdS–SNR samples. (a) Low magnification TEM image, (b) selected area electron diffraction pattern of a region marked by the dotted circle in (a), (c) HRTEM image showing lattice fringes, (d) Fourier filtered image of a region marked by the dotted box in (c), and (e) STEM-HAADF image; the inset shows the EDX line profile of a region marked by line 1. (f–j) EFTEM images taken from another core–shell nanostructure; (f) zero-loss map, (g) relative thickness map, (h) profile of the thickness map of a region marked by the rectangle in (g) showing a rod like structure, (i) chemical map of Nb (red) and (j) chemical map of Cd (green) indicating the locations of different atoms across the structure.

For detailed information of the Nb and Cd distribution in the core–shell nanorod, we have performed an elemental mapping using EFTEM as illustrated in Fig. 2(f)–(j). Energy filtered images were acquired using a contrast aperture of about 10 mrad to reduce aberrations (mostly chromatic). Chemical maps of the N-edges of Nb (34 eV) and Cd (67 eV) were obtained using a jump-ratio method by acquiring two images (one post-edge and one pre-edge), to extract the background, with an energy slit of 3 eV for Nb and 8 eV for Cd. The distribution of Cd throughout the nanorod in Fig. 2(j) reveals the formation of a thin layer around the sodium niobate nanorod.

The formation of nanorods possibly takes place only when the following steps occur in unison: (i) under highly basic reaction conditions, the precursor Nb(OC₂H₅)₅ hydrolyses to form Nb(OH)₆⁻ species which subsequently bind with Na⁺ species forming a Nb(OH)₆⁻Na⁺ complex,²² (ii) the viscous nature of ethylene glycol slows down the growth kinetics and (iii) ethylene glycol weakly binds with the OH⁻ groups of Nb(OH)₆⁻Na⁺, thus restricting the growth on the side facets.^23,24 In case of a hybrid sample, the thin layer of CdS grows onto pre-fabricated SNRs.

The band gap (E_g) which governs the light harvesting ability of semiconductors was determined for all samples from the reflectance data using the Kubelka–Munk method. The SNRs showed an E_g value in the order of 3.46 eV which is in agreement with the values reported in the literature,¹⁵ Fig. 3. As expected, a smaller E_g value of 3.37 eV was observed for BSN. The CdS–SNRs showed two E_g values of 3.6 and 2.59 eV which correspond to the sodium niobate and CdS, respectively. The appearance of two band gap values which are close to the band gap values of sodium niobate and cadmium sulfide further confirms the grafting of CdS on sodium niobate, and rules out the possibility of any doping/intercalation of CdS in sodium niobate.

Fig. 3 A plot of (αhν)² vs. hν for SNR, CdS–SNR and BSN samples.

To investigate the photoelectrochemical behavior, linear sweep voltammograms were recorded using an aqueous electrolyte solution containing 0.1 M Na₂S and 0.14 M Na₂SO₃ (pH ∼ 12.4) in a three electrode photoelectrochemical cell. The current density data obtained for CdS–SNRs, SNRs and BSN under irradiation with UV light (monochromatic light with a wavelength of 300 nm) and under darkness are shown in Fig. 4(A). These samples do not exhibit any significant anodic current under darkness. When CdS–SNRs are irradiated with UV light, a drastic enhancement in the anodic current density (J_an: 7.6 mA cm⁻² at 0.2 V vs. SHE) is observed, which is 3 times higher than that observed for SNRs (J_an: 2.5 mA cm⁻² at 0.2 V vs. SHE) and BSN (J_an: 0.6 mA cm⁻² at 0.2 V vs. SHE). The onset potential (E_on: the potential at which the current changes from cathodic to anodic), as determined from the linear sweep voltammogram, is much lower i.e. −0.7 V vs. SHE for CdS–SNRs than that observed for SNRs (E_on: ca. −0.6 V vs. SHE) and BSN (E_on: ca. −0.3 V vs. SHE). The thickness of the CdS shell is approximately 20 nm, which may allow the penetration of light (photons) to the NaNbO₃ shell. Therefore, the cumulative harvesting of high energy photons upon irradiation with 300 nm light of both the CdS shell and NaNbO₃ core may give rise to a higher number of photoexcitons and thus lead to a higher photocurrent in the case of CdS–SNRs than that observed with other photocatalysts. To further examine the effect of CdS grown on sodium niobate nanorods on photoelectrochemical behavior, the photocurrent was also measured for all samples under visible light using a solar simulator. The photocurrent density (current density under light − current density under darkness) data are shown for all samples in Fig. 4(B). The CdS–SNRs show a better photocurrent (J_an: 0.97 mA cm⁻² at 0.6 V vs. SHE), Fig. 4(B). The difference in the magnitude of the photocurrent density enhancement under UV and visible light for the CdS–SNR sample may be understood in terms of a better light harvesting efficiency by both CdS and sodium niobate in the UV region than that observed in the visible region by the CdS core (Fig. S4, ESI†). Despite the fact that they show wide E_g (UV region), the small photocurrent density observed for SNRs and BSN may partially be attributed to the harvesting of photons via surface states/defects (which lie in the band gap region) under visible light irradiation. The contribution of a small fraction of the UV irradiation, while using the solar simulator coupled with an AM 1.5G filter as a visible light source, cannot be ruled out. The solar simulator coupled with the AM 1.5G filter simulates the terrestrial solar spectrum on the ground when the sun is at a zenith angle of 48.2°. It includes both direct light from the sun and the diffuse light that is scattered by the atmosphere.

Fig. 4 Photo-catalytic response of CdS–SNR, SNR and BSN electrodes. (A) Current density recorded under darkness (grey lines) and monochromatic light with a wavelength of 300 nm (black lines), (B) photocurrent density recorded under visible light using a solar simulator; 5 mV s⁻¹, electrolyte containing 0.1 M Na₂S and 0.14 M Na₂SO₃.

In such a hetero-structured semiconductor, the relative band edge positions (conduction band, E_CB, and valence band, E_VB) of both semiconductors are critical and determine the efficacy of the exciton separation and their transfer across the interface. Therefore, the E_CB minima and E_VB maxima were determined for sodium niobate using the following equation.²⁷

where χ(A), χ(B), and χ(C) are the absolute electronegativity of atoms A, B, and C, respectively; E_c and E_g are the position of the conduction band and the band gap of the semiconductor, respectively; E₀ is the scale factor relating the reference electrode redox level to the AVS (E₀ = −4.5 eV for NHE). The schematic shows the relative position of E_CB minima and E_VB maxima as determined for the CdS–SNR hetero-structured semiconductor, Fig. 5. The E_CB value for sodium niobate (−0.76 V) is lower than the E_CB value of −1 V (as reported in the literature) for CdS. Such a downhill offset of the CdS–sodium niobate band edges seems to facilitate an efficient charge separation allowing the transfer of a photo-generated e⁻ from CdS CB to the sodium niobate CB (Fig. 5).

Fig. 5 Schematic representation showing the transfer of a photo-generated e⁻ from CdS to NaNbO₃ in CdS modified NaNbO₃ nanorods.

Fig. 6 Mott–Schottky plots for (A) CdS, (B) CdS–SNRs and (C) SNRs measured at 1 kHz.

The capacitance measurement of the electrode–electrolyte interface was conducted at 1 kHz frequency to determine the flat band potential (V_fb) and carrier density (N_D) of CdS, CdS–SNR and SNR electrodes using the following Mott–Schottky equation.²⁸

Here C and A are the interfacial capacitance and area, respectively, N_D is the number of donors, V is the applied voltage, k_B is Boltzmann's constant, T is the absolute temperature and e is the electronic charge. The slopes of the linear part of the curves in the Mott–Schottky plots are positive, indicating that CdS, CdS–SNRs and SNRs are n-type semiconductors. The V_fb values determined from the extrapolation for CdS, CdS–SNRs and SNRs are −0.76 V, −0.61 V and −0.64 V, respectively. The positive shift in the V_fb of CdS–SNRs relative to CdS and SNRs implies a decrease in band bending and thus facilitates electron transfer.²⁹ Further, the slope of the Mott–Schottky plot (Fig. 6) for CdS–SNR is higher, revealing its lower resistance for charge transport than the other samples.³⁰ The carrier density (N_D) values determined from the slope for CdS, CdS–SNRs and SNRs are 4.2 × 10²⁰ cm⁻³, 8 × 10²² cm⁻³ and 7.4 × 10²⁵ cm⁻³ respectively.

To further understand the e⁻ transfer dynamics, decay components were calculated from the time-resolved photoluminescence spectra (Fig. S5, ESI†). The best fitted parameters of decay dynamics for CdS–SNRs and SNRs are shown in Table 1. As can be clearly seen from the table, CdS–SNRs have a longer decay time than SNRs due to the effective charge transfer across the CdS–SNR interface. The relaxation or decay time mainly depends on the densities of initial and final states. Because of the surface passivation of the NaNbO₃ core by the CdS shell, there may be a decrease in the density of trapped states which leads to the decrease of fractional contribution f₂ but an increase in decay time τ₂. The longer average decay time (〈τ〉) of CdS–SNRs is due to a better charge separation in the staggered band gap of the CdS–SNR core–shell nanorods. These observations clearly show an efficient e⁻ transfer across the interface of CdS–SNRs.

Table 1 Emission decay components of the samples

Sample name	Decay time (ns)				Fractional contribution			Goodness of fit parameter (χ^2)
	τ1	τ2	τ3	〈τ〉	f1	f2	f3
SNR	0.33	1.89	9.95	6.45	0.45	0.42	0.13	1.117
CdS–SNR	0.60	2.35	15.20	8.06	0.55	0.38	0.07	1.070

---

It is worth noting that other niobate based photocatalysts reported in the literature show much lower photocurrent densities in the order of μA (<40 μA cm⁻²) under optimized conditions.^25,26 Such an improvement in the photocurrent for CdS–SNRs under UV and visible light may partially be attributed to the dual band gap values of CdS–SNRs which improve their overall light harvesting ability. The smooth lattice matched interface of CdS–sodium niobate as shown by HRTEM may also facilitate an efficient photo-generated e⁻ transfer across the interface, thereby, reducing the possibility of charge carrier loss by recombination.

In order to estimate the quantitative correlation of light harvesting to exciton (photo-generated e⁻ and h⁺) generation, the incident to photon conversion efficiency (IPCE) was calculated for all samples at wavelengths ranging from UV to visible solar radiation using the following equation.³¹

where I is the photocurrent density at 0 V vs. SHE, λ is the wavelength of the incident light and P_inc is the incident power density. The SNRs exhibit IPCE in the UV region. A small IPCE value in the order of 2.22% can be seen at 400 nm which increases to ca. 10% at 200 nm, Fig. 7. In the case of CdS–SNRs, the IPCE can be seen in both the visible and UV region and is many-fold higher than that observed for SNRs. It keeps increasing substantially from 6.55% (at 600 nm) to 23.91% (at 300 nm) and further increases steeply at 200 nm, Fig. 7. BSN and CdS show small IPCE values in the wavelength region 300 to 200 nm. The CdS shows lower IPCE (3.99% at 600 nm) than CdS–SNRs. These observations further corroborate the improved photo-generated charge carrier separation of the CdS–SNR electrode compared to the SNR, BSN and CdS electrodes.

Fig. 7 IPCE action spectra of the CdS, CdS–SNR, SNR and BSN electrodes.

Conclusions

In brief, a simple surfactant free synthetic method leads to the reproducible formation of sodium niobate and CdS modified sodium niobate nanorods. The modification of sodium niobate nanorods with narrow band gap CdS leads to a drastic improvement in the photocurrent density. Such an improvement in the photocurrent density appears to be due to the improved charge separation allowed by the staggered band gap of CdS–sodium niobate, smooth lattice matched interface of CdS–sodium niobate and the improved light harvesting ability. These results show the potential of such hetero-structured core–shell semiconductors, consisting of layered metal oxides, and visible light active semiconductors for efficient water splitting. They also serve as inspiration to fabricate hetero-structured photocatalysts based on layered metal oxides which provide ample room to tune their catalytic properties by intercalation and surface modifications.

Acknowledgements

Authors are grateful to MNRE (103/155/2009-NT) and CSIR (MLP-20, YSP – 04/2013), India for the financial assistance and Prof. B. K. Mishra, Director, IMMT Bhubaneswar for providing the support to undertake this work. Authors also thank Mr Sudhir K. Das and Dr M. Sarkar, NISER, Bhubaneswar for the time resolved photoluminescence measurements.

Notes and references

1. P. D. Tran, L. H. Wong, J. Barber and J. S. C. Loo, Energy Environ. Sci., 2012, 5, 5902.
2. Y. Li and J. Z. Zhang, Laser Photonics Rev., 2010, 4, 517.
3. R. van de Krol, Y. Liang and J. Schoonman, J. Mater. Chem., 2008, 18, 2311.
4. M. R. Allen, A. Thibert, E. M. Sabio, N. D. Browning, D. S. Larsen and F. E. Osterloh, Chem. Mater., 2010, 22, 1220.
5. W. Yao, C. Huang and J. Ye, Chem. Mater., 2010, 22, 1107.
6. K. Maeda and T. E. Mallouk, J. Mater. Chem., 2009, 19, 4813.
7. R. Ma, Y. Kobayashi, W. J. Youngblood and T. E. Mallouk, J. Mater. Chem., 2008, 18, 5982.
8. K. Maeda, M. Eguchi, W. J. Youngblood and T. E. Mallouk, Chem. Mater., 2009, 21, 3611.
9. Y. Miseki, H. Kato and A. Kudo, Energy Environ. Sci., 2009, 2, 306.
10. H. G. Kim, D. W. Hwang, J. Kim, Y. G. Kim and J. S. Lee, Chem. Commun., 1999, 1077.
11. T. Takata, Y. Furumi, K. Shinohara, A. Tanaka, M. Hara, J. N. Kondo and K. Domen, Chem. Mater., 1997, 9, 1063.
12. D. W. Hwang, H. G. Kim, J. Kim, K. Y. Cha, Y. G. Kim and J. S. Lee, J. Catal., 2000, 193, 40.
13. K. Saito and A. Kudo, Inorg. Chem., 2010, 49, 2017.
14. K. Saito, K. Koga and A. Kudo, Dalton Trans., 2011, 40, 3909.
15. H. Shi, X. Li, D. Wang, Y. Yuan, Z. Zou and J. Ye, Catal. Lett., 2009, 132, 205.
16. B. Z. Ska, E. B. Palen and R. J. Kalenczuk, J. Phys. Chem. Solids, 2011, 72, 117.
17. F. E. Sarac, C. Yilmaz, F. Y. Acar and U. Unal, RSC Adv., 2012, 2, 10182.
18. Y. Xie, G. Ali, S. H. Yoo and S. O. Cho, ACS Appl. Mater. Interfaces, 2010, 2, 2910.
19. N. Guijarro, T. L. Villarreal, Q. Shen, T. Toyoda and R. Gomez, J. Phys. Chem. C, 2010, 114, 21928.
20. N. Guijarro, Q. Shen, S. Gimenez, I. M. Sero, J. Bisquert, T. L. Villarreal, T. Toyoda and R. Gomez, J. Phys. Chem. C, 2010, 114, 22352.
21. D. R. Baker and P. V. Kamat, Adv. Funct. Mater., 2009, 19, 805.
22. M. Mazloum, S. Zanganeh, A. Kajbafvala, M. R. Shayeghl and S. K. Sadrnezhaad, Int. J. Eng., Trans. B, 2008, 21, 169.
23. G. J. de A. A. Soler-Illia, C. Sanchez, B. Lebeau and J. Patarin, Chem. Rev., 2002, 102, 4093.
24. C. Sanchez, J. Livage, M. Henry and F. Babonneau, J. Non-Cryst. Solids, 1988, 100, 65.
25. K. Yao, S. Nishimura, Y. Imai, H. Wang, T. Ma, E. Abe, H. Tateyama and A. Yamagishi, Langmuir, 2003, 19, 321; K. Yao, S. Nishimura, T. Ma, K. Okamoto, K. Inoue, E. Abe, H. Tateyama and A. Yamagishi, J. Electroanal. Chem., 2001, 510, 144; U. Unal, Y. Matsumoto, N. Tamoto, M. Koinuma, M. Machida and K. Izawa, J. Solid State Chem., 2006, 179, 33.
26. X.-T. Pian, B.-Z. Lin, Y.-L. Chen, J.-D. Kuang, K.-Z. Zhang and L.-M. Fu, J. Phys. Chem. C, 2011, 115, 6531.
27. J. Lv, T. Kako, Z. Li, Z. Zou and J. Ye, J. Phys. Chem. C, 2010, 114, 6157.
28. K. Gelderman, L. Lee and S. W. Donne, J. Chem. Educ., 2007, 84, 685.
29. Z. Zhang, Y. Yu and P. Wang, ACS Appl. Mater. Interfaces, 2012, 4, 990.
30. K. H. Yoon, C. W. Shin and D. H. Kang, J. Appl. Phys., 1997, 81, 7024.
31. J. Gong, C. Lin, M. Ye and Y. Lai, Chem. Commun., 2011, 47, 2598.

---

Footnote

† Electronic supplementary information (ESI) available: XRD pattern for SNRs, CdS–SNRs and BSN, SEM image of BSN, TEM image of CdS, light harvesting efficiency and time-resolved photoluminescence spectra of SNRs and CdS–SNRs. See DOI: 10.1039/c3ra47024e

---