1. Introduction With the increasing demand for clean and renewable energy to solve problems related to the emission of greenhouse gases and the depletion of fossil fuels, hydrogen, with high gravimetric energy density, has been considered as an attractive fuel for the future. However, even today, hydrogen is mainly produced from fossil fuels, which aggravates the emission of greenhouse gases inevitably [1]. Taking into account the solution of energy and environmental issues, many research groups have focused on finding ways to produce H2 from water in the sunlight, such as reforming of biomass, electrolysis of water by solar cells, and especially photoc water splitting [2]. Since the first report on the photocatalytic water splitting by Fujishima and Honda in the early 1970s [3], hydrogen produced by photocatalytic water splitting using TiO2-based photocatalysts has been considered as a promising way to produce hydrogen environmental-friendly and cost-effective [4e6]. It is widely known that the photocatalytic activity of pure TiO2 is restricted by its large electronic band gap (anatase, 3.2 eV) and the recombination of photoexcited electronehole pairs [7,8]. To address these drawbacks, doping nonmetal elements into TiO2, such as N [9e16], C [17], S [18] and F [19], is found to be an effective method to enhance the activity of TiO2 photocatalyst and thereby is widely adopted. On the other hand, it is also well-known that the performance of TiO2 is not only governed by its composition and crystal structure, but also by its surface properties [20e22]. Recently, many studies have shown that TiO2 crystals with a large percentage of high energy facets exhibit excellent photocatalytic activity [23e27]. Therefore, it is believed that the cooperation of nonmetal doping and preferred orientation can lead to effective hydrogen production when TiO2 films are used as the catalyst. Yet, very few studies [28,29] have been carried out in this area. In this communication, N-doped TiO2 films with preferred (211) orientation were deposited by RF magnetron sputtering. And the experimental results showed that not only the N doping, but also the preferred (211) orientation, could improve the hydrogen production ability of TiO2 film significantly. 2. Experimental N-doped TiO2 films were synthesized on quartz glass substrates (2 cm   4 cm) at a fixed temperature of 450 C. The target was a metallic plate of Ti (99.999% pure), 2-inch in diameter. The target-substrate distance was fixed at 70 mm and the base vacuum was 5.8  10
   3 Pa. The working gas was a mixture of N2 (99.999% pure), O2 (99.99% pure) and Ar (99.999% pure). Before the working gas was introduced into the chamber, the target was pre-sputtered by argon plasma for 30 min. Then, 3 mass flow controllers were employed to fix the flow rate of O2, Ar and N2 at 3.9 sccm (standard cubic centimeter per minute), 40 sccm and 0e8 sccm, respectively. The samples deposited with N2 flow rate of 0, 2, 5, and 8 sccm were named as TN-0, TN-2, TN-5, and TN-8, respectively. For all the samples, the total working pressure was set at 0.3 Pa and the RF power was fixed at 130 W. The whole deposition was carried out for 4 h. Pure TiO2 film was deposited as reference under the same condition without N2. The chemical states and compositions of the undoped and N-doped TiO2 films were identified by X-ray photoelectron spectroscopy (XPS, ESCALAB 250, Thermo Scientific, USA). The crystalline structure of the deposited N-doped TiO2 films was determined by X-ray diffraction (XRD, X0 Pert Pro MPD, Panalytical, Holland). The diffractometer was operated at 40 kV and 40 mA with a Co Ka (lCo Ka ¼ 0.179 nm). The hydrogen production was tested by the photocatalytic water splitting testing system (CEL-SPH2N, AULTT, China), using a 300W Xe lamp (CEL-HXF300, AULTT, China) as irradiation source. The film-lamp distance was fixed at 12 cm. The photocatalytic reaction was carried out in a quartz cell containing 100 mL methanol aqueous solution (CH3OH: H2O ¼ 1:10 v/v). The amount of H2 was analyzed by a gas chromatograph (SP7800, AULTT, China). 3. Results and discussion 3.1. Crystalline structure and chemical composition In order to determine the chemical states and compositions of the undoped and N-doped TiO2 films, XPS experiments were carried out. All the films exhibited Ti-2p3/2 peaks at 458.5 eV and O-1s peaks at 529.7 eV, data not show here (as shown in Fig. S1 in the Supporting Information). Fig. 1 shows the highresolution N-1s XPS spectra of the films and the corresponding fitting curves. For the undoped sample (TN-0), only a weak peak at 399.8 eV was observed, while for the N-doped samples, a pair of N-1s features located at 396.5 eV and 399.8 eV were detected respectively. In general, the peak at 399.8 eV is assigned to either chemisorbed molecular N2 on the surface or interstitial molecular N2 [30], while the peak at 396.5 eV is attributed to atomic b-N states (i.e., OeTieN bonds) [31,32]. Therefore, the peak at 396.5 eV for the N-doped samples indicates the substitutional incorporation of N in N-doped TiO2 films [33]. From the estimate for the integrated intensities of the N-1s peak at 396.5 eV, it is found that the atomic concentration of the incorporated N was 4.76 at.% for TN-2, 4.91 at.% for TN-5 and 4.33 at.% for TN-8. The little difference in nitrogen concentration for all the N-doped samples implies that the incorporation of N in the films was not promoted by the increasing N2 flow rate. Fig. 2 illustrates the XRD patterns of undoped and N-doped TiO2 films. All the samples are of pure anatase phase (JCPDS 

Fig. 1 e X-ray photoelectron spectral collected from Ndoped TiO2 and pure TiO2 films