Modulating the interaction between gold and TiO2 nanowires for enhanced solar driven photoelectrocatalytic hydrogen generation

Please note that technical editing may introduce minor changes to the text and/or graphics, which may alter content. The journal’s standard Terms & Conditions and the Ethical guidelines still apply. In no event shall the Royal Society of Chemistry be held responsible for any errors or omissions in this Accepted Manuscript or any consequences arising from the use of any information it contains. Accepted Manuscript


Introduction
Semiconductor photocatalysis and photoelectrochemistry, particularly involving TiO 2 , is an influential field that plays an important role on environmental remediation and energy conversion applications.The striking features of TiO 2 including its high chemical stability in aqueous media, high photoactivity, earth abundance and environmentally benign nature strongly encourage the use of this material as a potential electron acceptor in light driven devices operating under solar radiation. 1,2The primary report on photoelectrocatalytic (PEC) water oxidation on TiO 2 by Fujishima and Honda 3 has promoted a huge body literature on the design of nanostructured photocatalytic materials for realizing the solar driven PEC water oxidation process. 4,5Moreover, the photogenerated holes at the valence band of TiO 2 energetically favour the oxidation process of chemical components (like pollutants, etc.).This process can be understood as a renaissance of natural photosynthesis utilizing manmade materials.Nonetheless, the short wavelength cut-off property of TiO 2 permits accessing only a narrow segment (B5%) of the solar spectrum; resulting in low PEC efficiency. 6onversely, the VB edge position (B3.2V vs. NHE) of TiO 2 meets the energy requirement for successful PEC water oxidation, where most of the visible light band gap materials such as CdS, CdSe, GaAs, and GaP fall short. 7,8Therefore, developing strategies to amplify the light harvesting efficiency of TiO 2 (greater than 5%) without sacrificing its PEC water oxidation potential is strongly in demand to further improve the PEC performance of TiO 2 .
0][11] Due to LSPR, Au NPs can effectively harvest the electromagnetic energy of the incident light and concentrate it into ''plasmonic hotspots'', which in turn can maximize the local field intensity by a factor of 10 6 . 12otably, one dimensional (1-D) TiO 2 nanostructures such as nanowires, nanotubes, nanofibers, etc. afford ample room to accommodate the nanoparticles compared to their bulk counterparts. 13Moreover, the rapid electron transport in the 1-D TiO 2 crystallite framework is highly beneficial for the efficient charge collection at the relevant interfaces. 14,158][19][20][21] This indicates that the interfacial interaction between Au NPs and TiO 2 plays a key role in achieving optimum PEC performance.
So far, a number of strategies have been documented in the literature for anchoring the Au NPs on TiO 2 surfaces.For instance, Au NPs were coated on TiO 2 via physical evaporation or sputtering resulting in the formation of inhomogenous NP coating on the metal oxide host surface. 22,235][26][27][28] Unfortunately, such deposition techniques affect the supporting semiconductors due to the multiple post-treatment steps.Similarly, controlling the size, shape and distribution of Au NPs using such approaches still remains challenging.Recently, we have demonstrated a straightforward method of anchoring 4-dimethylaminopyridine (DMAP) capped Au NPs on TiO 2 hollow nanowires, which in turn resulted in enhanced photocatalytic performance towards dye degradation. 29n the present study, we investigate the underlying mechanism behind the single-step anchoring of DMAP capped Au NPs on TiO 2 nanowires and their respective interactions were studied using theoretical simulations based on first principle theory.The interaction between DMAP capped Au NPs and TiO 2 nanowires was modulated via nitrogen doping TiO 2 .The influence of nitrogen doping and DMAP capped Au NPs anchoring on TiO 2 band modification was elucidated.The finite-difference time domain (FDTD) method was used to quantitatively analyse the electromagnetic field enhancement in the TiO 2 -Au hybrid photocatalytic system.Our theoretical and experimental results demonstrate the synergistic effect of Au NPs (visible light scattering) and nitrogen doping (band gap narrowing effect) on the PEC water oxidation process and the implications of these findings are discussed.

Materials
All materials were received from Sigma-Aldrich and used as received.

Fabrication of TiO 2 nanowire arrays
Vertically aligned TiO 2 hollow nanowires onto FTO substrates were grown using ZnO nanorod array templates.First step, ZnO film of 200 nm was deposited on the FTO substrate using radio frequency magnetron sputtering.Following that, the ZnO seed layer coated FTO substrate was immersed into aqueous solution of 0.025 M zinc nitrate hexahydrate and 0.025 M hexamethylenetetramine and the sample was kept at 85 1C for 10 h for the growth of vertically aligned ZnO nanorod arrays.After ZnO nanorod growth, the substrates were rinsed with deionized water.
In the second step, the TiO 2 layer will be grown on ZnO nanorod arrays as follows: the resultant ZnO NR array template was kept in aqueous solution of 0.075 M ammonium hexafluorotitanate and 0.2 M boric acid.This chemical bath deposition results in the thin layer of TiO 2 coating ZnO nanorods.Subsequently, the TiO 2 coated ZnO electrode is immersed in a 0.5 M boric acid solution for 1 h and the ZnO template is removed and rinsed with deionized water.Finally, vertically aligned TiO 2 NW arrays are obtained.The TiO 2 NW samples were calcined at 500 1C for 0.5 h under an Ar atmosphere.
For nitridation onto TiO 2 NWs, the electrodes were transferred to a chemical vapour deposition (CVD) chamber.The nitridation process was carried out at 500 1C with H 2 and NH 3 with flow rates of 50-200 standard cubic centimetre per minute (sccm) and 100-300 sccm, respectively.

Synthesis of Au NPs
DMAP-capped Au NPs were synthesized using the phase transfer procedure. 30Briefly, 0.030 M aqueous solution of HAuCl 4 Á3H 2 O was added to 0.025 M tetraoctylammonium bromide (TOAB) in toluene.Then, 0.4 M aqueous NaBH 4 was added drop-wise to the mixture with stirring, causing an immediate reduction to occur.After 24 h, the two phases were separated and the toluene phase was subsequently washed with 0.1 M H 2 SO 4 , 0.1 M NaOH, and H 2 O (three times), and then dried over anhydrous Na 2 SO 4 .An equal volume of 0.1 M aqueous solution of DMAP was then added.The phase transfer is clearly visible as the dark pink coloured solution transfers from toluene to water due to the addition of DMAP and was completed within 1 h.Assuming a 100% efficient reduction of gold chloride and no losses during transfer and washing steps, the particle size yields a particle concentration of approximately 6.8 Â 10 À7 M in the stock solution arising from the nanoparticle synthesis.

Assembly of Au NPs on TiO 2 nanowire arrays
Au NP coating on TiO 2 and N-TiO 2 NW arrays was obtained by immersing the TiO 2 electrodes in positively charged, water soluble DMAP-capped Au NP solution (3.4 Â 10 À7 M) for 5 min.The positively charged DMAP-capped Au NPs were attracted towards the negatively charged surface of TiO 2 NWs and N-TiO 2 NWs, resulting in the facile formation of electrostatically assembled TiO 2 -Au and N-TiO 2 -Au hybrid photocatalytic electrodes.After coating Au NPs on these electrodes DMAP molecules were removed by the sintering process.

Theoretical calculations
The energy calculations were performed using the PW91 method generalized gradient approximation (GGA) and the plane wave model using CASTEP program.The supercell of 20 Å Â20 Å Â c was used for the calculation, which represents the nanowire length along the tube axis as c.All atoms were described using Vanderbilt ultrasoft pseudopotentials and a cut off energy of 240 eV where a set of k-points was used to expand the electronic wave function based on the Monkhorst-Pack scheme within 5.0 Â 10 À5 eV per atom of total energy convergence.Electron density was investigated with optimized geometries.The binding energies of atoms on the nanotubes were calculated by eqn ( 1) (1)

Structural and optical characterization
The surface of various TiO 2 nanowire arrays was characterized using a field emission scanning electron microscope (FE-SEM, JEM-3100F, Jeol, Tokyo, Japan) and a field emission transmission electron microscope (FE-TEM, JSM 7600F, JEOL, Tokyo, Japan).The chemical environment of pure and N doped TiO 2 electrodes was analyzed by X-ray photoelectron spectroscopy (XPS) using an angular resolved electron analyzer with a monochromated Al Ka source (Theta Probe, Thermo Fisher Scientific).
The optical diffuse reflectance spectra of the electrodes were recorded in the range of 350-900 nm using a V670 JASCO UV-Vis spectrophotometer.The absorbance of the electrodes was estimated from diffuse reflectance measurements (R) directly using the Kubelka-Munk relation, F.R. = (1 À R) 2 /2R using in built software from a JASCO UV-Vis spectrometer.We have assumed that the scattering term is similar for photoelectrodes with and without Au loading.

(Photo)-electrochemical characterization
The (photo)-electrochemical analysis was done using a three electrode configuration.The as-prepared TiO 2 , TiO 2 -Au, N-TiO 2 and N-TiO 2 -Au electrodes (electrode area 1 cm 2 ) were used as the working electrode, Ag/AgCl as the reference and Pt foil as the counter electrode.0.5 M of Na 2 SO 4 (Sigma Aldrich) (pH = 6.1) was used as the electrolyte for all PEC measurements without any additional additives.Cyclic voltammograms were recorded using an advanced potentiostat (PGSTAT-30 from Autolab) with a scanning rate of 50 mV s À1 .The photocurrent measurements were recorded using a solar simulator with a 300 W xenon arclamp (Hayashi-LA 251-Xe).The light intensity was calibrated using a silicon photodiode (100 mW cm À2 ).The electrolyte was bubbled with nitrogen gas for 30 min to avoid the presence of oxygen (electron acceptor) in the solution.The IPCE measurements were carried out by employing a 300 W Xe lamp coupled to a computer-controlled monochromator; the photoelectrode was polarized at the desired voltage (1.6 V vs. RHE) using a Gamry potentiostat, and the photocurrent was measured using an optical power meter 70 310 from Oriel Instruments.A Si photodiode was used to measure the light intensity to calibrate the system.The output gas samples were collected from the head space of a sealed PEC chamber using an air-tight gas syringe through the manual sampling port in the top of the chamber (flexible cork made of Teflon) and further subjected to gas chromatographic analysis to evaluate the constituents of the gas products.

Theoretical calculations
To get insights into the binding mechanism of positively charged DMAP-capped Au NPs on TiO 2 NWs and N-TiO 2 NWs, it is important to estimate the binding energies in the hybrid systems.First principles based simulations provide direct evidence of interaction energies between DMAP-Au NPs and the TiO 2 semiconductor, which can be used to detect the optimal binding sites and to predict the maximal affinity that a nanoparticle could attain from them.The binding characteristics in TiO 2 -Au and N-TiO 2 -Au hybrid photocatalytic systems were analyzed using first principles theory.Firstly, the binding energies between molecular DMAP and Au were estimated.It was found that the DMAP molecule strongly binds to gold with À14.01 kcal mol À1 binding energy, which could be ascribed to the high binding affinity of Au atoms towards the pyridine moieties in DMAP.
2][33] Moreover, it was realized that the deposition of metal nanoparticles (Ag, Au and Pt) onto selective {101} facets of anatase TiO 2 crystals with different percentages of exposed {001} and {101} facets can effectively enhance the photocatalytic activity of TiO 2 in both photoreduction and photo-oxidation processes. 34,35In this line, we examined the binding nature of DMAP-Au NPs on {101} and {001} facets of TiO 2 .Fig. 1 shows the optimal binding sites of DMAP-Au NPs on (a) {001} facets and (b) {101} facets of TiO 2 .The {101} facet of TiO 2 shows higher binding energy (À8.282 kcal mol À1 ) towards DMAP-Au NP anchoring compared to the {001} facets of TiO 2 (À6.48 kcal mol À1 ), illustrating the higher affinity of {101} surfaces towards DMAP-Au NP anchoring.Fig. 1c shows the density of the electron cloud at the TiO 2 -Au hybrid electrode.
The reported excellent catalytic activity and band structure modifications of TiO 2 upon nitrogen doping [36][37][38][39] inspired us to investigate the effect of nitrogen doping on DMAP-Au NP binding.Fig. 1(d-f) show the binding orientations and density of electron cloud distribution in N-TiO 2 -Au hybrid electrodes.As anticipated, higher binding energies (À8.86 kcal mol À1 ) were observed for DMAP-Au NPs anchoring on N-TiO 2 than on pristine TiO 2 NWs (À6.48 kcal mol À1 ) at the {001} surface.This could be ascribed to the higher binding affinity of pyridines and Au towards nitrogen.Strikingly, the binding energies were found to be remarkably enhanced at the {101} surface of N-TiO 2 NWs.The N-TiO 2 -Au hybrid system shows a binding energy of À35.56 kcal mol À1 at the {101} surface, which is nearly four times higher than that observed for the {001} plane of N-TiO 2 NWs.This strongly suggests that the Au NPs can be effectively coated onto N-TiO 2 by electrostatic attraction.
On the other hand, in view of improving the light harvesting efficiency of TiO 2 , Au NP decoration found to improve the visible light activity of hybrid materials owing to their strong scattering effect and enhanced absorption at around 520 nm due to the LSPR band.To understand the origin and the extent of improved visible light activity in TiO 2 -Au hybrid systems, the light scattering effect of Au NPs at TiO 2 NWs was further analyzed using the finite-difference time-domain (FDTD).Fig. 2 shows the electric field (E-field) distribution in the perpendicular direction, across the Au NPs at the TiO 2 -NW surface obtained from FDTD simulations.The permittivity of gold was analyzed using the Lorenz-Drude dispersive model. 40r ðoÞ ¼ e

Experimental results
From the above discussion, it is inferred that the Au NP decoration at TiO 2 NWs could yield high PEC energy conversion efficiency, mainly through light scattering and co-catalyst effects.Moreover, nitrogen doping was found to enhance the interaction between Au NPs and TiO 2 as well as the visible light activity of TiO 2 by the band gap narrowing effect. 39In order to understand the role of Au NP decoration and nitrogen doping,  we have designed and fabricated Au NP decorated TiO 2 NW and N-TiO 2 NW arrays onto fluorinated tin oxide (FTO) substrates.Fig. 3 shows the electron microscopy images of nitrided TiO 2 (N-TiO 2 ) NW, Au NP coated TiO 2 NW (TiO 2 -Au) and N-TiO 2 NW (N-TiO 2 -Au) arrays, respectively.Scanning electron microscopy (SEM) and transmission electron microscopy (TEM) images of N-TiO 2 NWs clearly reveal their vertically aligned geometry and 1-D structure (Fig. 3a).The N-TiO 2 NWs have a length of B5 mm, an inner free space of B200 nm and a shell thickness of B50 nm, respectively (Fig. 3b).
From Fig. 3c and d, a thin amorphous layer is observed on TiO 2 .This layer represents the formation of TiN/TiO x N y on the TiO 2 NW surface during the nitridation process.Although it was difficult to obtain lattice resolved TEM images at the surface of N-TiO 2 NWs, the X-ray diffraction (XRD) and SAED patterns support that N-TiO 2 NWs maintain their polycrystalline anatase phase (see ESI, † Fig. S1).There are no noticeable changes in the geometry of the NWs after Au NP anchoring (Fig. 3f and j).Au NPs with an average diameter of B5 nm were observed in the high magnification TEM images (Fig. 3i) (also see ESI, † Fig. S2).These images (Fig. 3h and l) confirm the higher loading of Au NPs on N-TiO 2 NWs compared to pristine TiO 2 NWs.
The effect of nitrogen doping on TiO 2 NWs was studied by X ray photoelectron spectroscopy.Fig. 4(a) shows the core spectra of Ti 2p in pristine TiO 2 NWs and N-TiO 2 NWs.The shoulder peaks at 459.4 eV and 465.1 eV correspond to Ti 2p 3/2 and Ti 2p 1/2 , respectively. 41In N-TiO 2 NWs, these peaks were shifted toward lower binding energy indicating the successful nitrogen doping at the TiO 2 NWs.Upon further examining the N 1s core spectra (Fig. 4b), the peaks at 396.8, 398.8, and 400.6 eV provide a clear picture of the nitrogen environment at TiO 2 NWs.The peak at 396.8 eV is attributed to the existence of atomic b-N in the TiO 2 matrix and the peak at 398.8 eV implies the substitution of O 2 À by nitrogen ions and thus, results in the formation of N-Ti-O. 42,43This confirms the existence of Ti 3+ due to the incorporation of nitrogen atoms into oxygen lattice sites.The weaker peak at 400.6 eV indicates the interstitial doping which induces an additional impurity state above the valence band of TiO 2 . 44The atomic ratio of nitrogen doping carriers was estimated from Fig. 4b and found to be 5.8 wt%.
The broad features of the O 1s peak in Fig. 4c were deconvoluted using the Gaussian fit (Fig. 4c, inset).Three distinct peaks were observed at 530.9, 532.2 and 533.2 eV representing the lattice oxygen, surface hydroxyl oxygen and surface adsorbed oxygen, respectively.The atomic weight of lattice oxygen at TiO 2 NWs was estimated to be 51.5%, which was reduced to 47.6% upon nitrogen doping.This indicates the possibility of nitrogen atoms occupying the oxygen vacancies.The influence of nitrogen doping on the work function of TiO 2 NWs was examined using ultraviolet photoelectron spectroscopy (see ESI, † Fig. S3).The valence band maximum (VBM) position of TiO 2 NWs at B3.26 eV was shifted to B2.69 eV upon nitrogen doping.This indicates that nitrogen doping carriers  are creating sub-bands or defects above the VB of TiO 2 as discussed in the XPS results.
In order to study the contribution of DMAP-Au NPs towards the visible light activity of TiO 2 -Au and N-TiO 2 -Au hybrid photocatalysts, diffuse reflectance measurements were performed.The results are shown in Fig. 5a.The Kubelka-Munk absorbance contribution of TiO 2 NWs and N-TiO 2 NWs was subtracted from the absorbance spectrum of TiO 2 -Au and N-TiO 2 -Au hybrids, respectively (Fig. 5b).Both spectra confirm the enhanced scattering induced by the presence of Au NPs, being the LSPR effect negligible for these specimens.Although, Au NP coating at TiO 2 and N-TiO 2 was carried out under identical conditions, N-TiO 2 -Au hybrid electrodes showed higher absorbance, indicating higher loading of Au NPs in good correspondence with the TEM measurements (Fig. 2h and l) and the Kubelka-Munk function (F.R.) (inset of Fig. 5).This is attributed to the higher binding energy observed at the N-TiO 2 surfaces.The broad absorption nature of N-TiO 2 -Au may be originated from the light scattering effect which amplifies the absorbance of TiO 2 . 45o test the photoelectrochemical (PEC) water oxidation performance, we measured J-V characteristics of these electrodes under the dark and illumination conditions.From Fig. 6a, it was found that the TiO 2 -Au hybrid electrode showed four times higher photocurrent B0.4 mA cm À2 than the pristine TiO 2 NW electrode (0.11 mA cm À2 ).This photocurrent enhancement may be due to the combined effect of higher optical absorption due to the scattering and co-catalytic effect of Au NPs in water oxidation through the formation of the Au-TiO 2 Schottky junction. 46,47In addition, it may facilitate the charge separation at TiO 2 /electrolyte interfaces and thus reduces the recombination of electrons through surface states of TiO 2 . 48In the case of nitrogen doping at TiO 2 NWs, the photocurrent density was slightly enhanced (0.22 mA cm À2 ), which might be ascribed to the promotion of visible light activity of TiO 2 through the band gap narrowing effect (see ESI, † S3 and S4).
In striking contrast, the photocurrent in the N-TiO 2 -Au hybrid system was only slightly higher than that in the N-TiO 2 NW system.This might be due to excessive Au NP decoration (from Fig. 3l), which may result in forward light scattering blocking the light photons reaching the N-TiO 2 NW surface and also reduce the interfacial contact between N-TiO 2 and FTO electrodes, thus hindering the photoholes participating in the water oxidation process. 18This implies that there exists a tradeoff between the enhanced light scattering due to Au NP decoration and visible light photon reception at N-TiO 2 NWs, which necessitates the optimization of Au NP loading onto N-TiO 2 NWs.We have recently reported that the photoelectrochemical performance of TiO 2 -Au nanocomposites has a volcano dependence with Au loading. 28nterestingly, the Au NP decorated TiO 2 and N-TiO 2 electrodes using low concentration stock solution (0.34 Â 10 À7 M), here referred to as TiO 2 -Au (low) and N-TiO 2 -Au (low), respectively, lead to higher photocurrent compared to the high Au concentration samples, as observed in Fig. 6.This photocurrent enhancement strongly suggests minimizing the optical blocking effect through controlling the Au NP loading on N-TiO 2 .
Furthermore, the stability of electrodes in photocurrent generation is examined with chronoamperometric curves (see ESI, † S5).The significant decrease of the photocurrent in both TiO 2 and TiO 2 -Au electrodes may be attributed to the current leakage at TiO 2 /FTO interfaces to the electrolyte through the naked FTO surface (uncovered TiO 2 area).Another plausible reason may arise from Au stability under long time photo-irradiation.However it is not clear at this moment and further research on this topic is needed.It is anticipated that inserting a compact TiO 2 blocking layer between TiO 2 and FTO layers will hinder the electron flow from the charge collector to the electrolyte.
In order to further corroborate the absence of the plasmonic effect at Au coated TiO 2 electrodes, we recorded IPCE spectra (see ESI, † Fig. S6).There is no photocurrent peak at around 550 nm in the TiO 2 -Au electrode as observed in optical absorption spectra (Fig. 5b).This implies that there is no electron injection from Au nanoparticles to TiO 2 by the plasmonic effect.Instead, Au NPs are promoting the optical absorption of TiO 2 through the scattering effect.In addition, Au NPs play a critical role as a co-catalyst in the water oxidation process as shown in Fig. S4 (ESI †).The photoholes generated at the  Aiming at the technological exploitation of the photocurrent enhancement due to nitrogen doping and loading of Au nanoparticles, we carried out the deposition of these structures on flexible stainless steel (SS) substrates.The flexible photoelectrodes are highly promising in solar-to hydrogen fuel production owing to their lower weight and cost, compared to FTO, together with more versatile mechanical properties.However, the TiO 2 nanowire fabrication onto flexible SS substrates remains challenging.A similar deposition procedure of TiO 2 on FTO substrates explained in the Experimental section was repeated for fabricating TiO 2 NWs on flexible SS substrates.We adopt the optimized Au NP concentration condition (0.34 Â 10 À7 M) from Fig. 6, and reconstructed the TiO 2 -Au hybrid systems onto flexible stainless steel (SS) substrates instead of FTO substrates (the sample photo is presented in the inset of Fig. 7a).
The J-V characteristics of Au-TiO 2 -SS and N-TiO 2 -Au-SS electrodes were tested in the 0.5 M Na 2 SO 4 aqueous electrolyte with a three electrode PEC cell step up.The J-V characteristic of N-TiO 2 -Au-SS is displayed in Fig. 7a.The higher photocurrent density achieved in N-TiO 2 -Au-SS compared to the TiO 2 -Au-SS electrode agrees well with the trends presented in Fig. 6.The scattering effect by Au NPs, the band gap narrowing effect by N doping and catalysis at optimal concentration lead to high photocurrent generation at the N-TiO 2 -Au-SS sample.A small onset potential shift occurred in this sample towards the negative potential region, also in good agreement with the results shown in Fig. 6.We believe that this cathodic shift can be related to enhanced charge collection at the contact for the N-doped TiO 2 .
The gas evolution from the PEC water oxidation experiment was collected for one hour under light illumination at 1 V vs. RHE potential, and analysed using gas chromatography (Fig. 7b).It is observed that TiO 2 -Au-SS hybrid electrodes could generate hydrogen of about 120 mmol h À1 cm À2 .Interestingly, the hydrogen generated at N-TiO 2 -Au-SS hybrid electrodes was about 270 mmol h À1 cm À2 .The described bendable architecture of N-TiO 2 -Au on SS substrates (inset of Fig. 7a) with light weight and low cost can be beneficial for large scale flexible PEC fuel generation. 49,50Another advantage of both side conducting nature of SS substrates enables us to design an artificial leaf based (back side coated with the electrocatalyst) on monolithic wireless type solar fuel cells. 51

Conclusions
Hybrid photocatalytic systems developed by anchoring DMAP-Au NPs on N-TiO 2 nanowires exhibit interesting optical and photocatalytic properties for water oxidation.Nitrogen doping of TiO 2 both maximizes the DMAP-Au NP loading and improves the optical absorption due to the band narrowing effect.DMAP-Au NP anchoring amplifies the visible light activity of TiO 2 and N-TiO 2 via a light scattering effect and also improves the charge recombination at the electrode/electrolyte interface.After optimization of the Au loading on the nanostructured photoelectrodes, N-doped TiO 2 leads to enhanced photocurrent generation and slightly more favourable onset potential on both FTO and SS substrates.Hydrogen generation measured on films deposited on flexible SS substrates was significantly higher on N-doped TiO 2 nanostructures.

a
Photocatalysis International Research Center, Research Institute for Science &

2 )
where e r (o) is the relative permittivity at infinity frequency, G m is the strength of each resonance term, O m is the plasma frequency, o and o m are the angular and resonant frequency, respectively, and G m is the damping factor or collision frequency.Fig.2(a) and (b), clearly illustrate the electric field generation around the Au NPs.The electric field extension varies with Au NP loading (spacing) on TiO 2 .The visible light scattering centres around Au NPs transformed the light reception at TiO 2 .However, higher loadings of Au NPs on TiO 2 found to scatter the major portion of input light, which may reduce the light transmission reaching the TiO 2 surface.Thus, it is anticipated that the TiO 2 -Au hybrid system benefits not only from improved LSPR band facilitated absorption at 520 nm, but also from the light scattering effect of Au NPs, which could effectively improve the light absorption properties of TiO 2 itself.

Fig. 1
Fig. 1 The optimal orientations of TiO 2 (a and b) and N-TiO 2 (d and e) surfaces illustrating the binding mechanism of DMAP-Au NPs at {001} and {101} planes, respectively.The electron density distribution estimated at TiO 2 -Au (c) and N-TiO 2 -Au (f) electrodes.

Fig. 2
Fig. 2 Spatial distribution of electric-field energy density of the TiO 2 -Au hybrid photocatalytic system under different spacing conditions: (a) with a few nm spacings and (b) Au nanoparticles in contact (the y axis indicates the distance in nm).

Fig. 5
Fig. 5 (a) Absorbance spectra of TiO 2 and N-TiO 2 electrodes in the presence and the absence of Au NPs; (b) Kubelka-Munk relation of Au NPs decorated at TiO 2 NWs and N-TiO 2 NWs.

Fig. 6 J
Fig. 6 J-V results of PEC water splitting using different photoanode electrodes.Note that two different Au concentrations were used for decorating working electrodes: (low -0.34 Â 10 À7 M; high -3.4Â 10 À7 M).