Significantly Enhanced Visible Light ... - ACS Publications

21 downloads 0 Views 6MB Size Report
Jun 8, 2015 - California Nanosystems Institute, University of California, Los Angeles, California 90095, United States. §. Department of Physics, Hubei ...
Letter pubs.acs.org/NanoLett

Significantly Enhanced Visible Light Photoelectrochemical Activity in TiO2 Nanowire Arrays by Nitrogen Implantation Gongming Wang,†,‡,¶ Xiangheng Xiao,†,§,¶ Wenqing Li,§ Zhaoyang Lin,† Zipeng Zhao,∥ Chi Chen,⊥ Chen Wang,‡,∥ Yongjia Li,∥ Xiaoqing Huang,∥ Ling Miao,⊥ Changzhong Jiang,*,§ Yu Huang,‡,∥ and Xiangfeng Duan*,†,‡ †

Department of Chemistry and Biochemistry, University of California, Los Angeles, California 90095, United States California Nanosystems Institute, University of California, Los Angeles, California 90095, United States § Department of Physics, Hubei Nuclear Solid Physics Key Laboratory and Center for Ion Beam Application, Wuhan University, Wuhan 430072, People’s Republic of China ∥ Department of Materials Science and Engineering, University of California, Los Angeles, California 90095, United States ⊥ School of Optical and Electronic Information, Huazhong University of Science and Technology, Wuhan 430074, People’s Republic of China ‡

S Supporting Information *

ABSTRACT: Titanium oxide (TiO2) represents one of most widely studied materials for photoelectrochemical (PEC) water splitting but is severely limited by its poor efficiency in the visible light range. Here, we report a significant enhancement of visible light photoactivity in nitrogen-implanted TiO2 (N-TiO2) nanowire arrays. Our systematic studies show that a post-implantation thermal annealing treatment can selectively enrich the substitutional nitrogen dopants, which is essential for activating the nitrogen implanted TiO2 to achieve greatly enhanced visible light photoactivity. An incident photon to electron conversion efficiency (IPCE) of ∼10% is achieved at 450 nm in N-TiO2 without any other cocatalyst, far exceeding that in pristine TiO2 nanowires (∼0.2%). The integration of oxygen evolution reaction (OER) cocatalyst with N-TiO2 can further increase the IPCE at 450 nm to ∼17% and deliver an unprecedented overall photocurrent density of 1.9 mA/cm2, by integrating the IPCE spectrum with standard AM 1.5G solar spectrum. Systematic photoelectrochemical and electrochemical studies demonstrated that the enhanced PEC performance can be attributed to the significantly improved visible light absorption and more efficient charge separation. Our studies demonstrate the implantation approach can be used to reliably dope TiO2 to achieve the best performed N-TiO2 photoelectrodes to date and may be extended to fundamentally modify other semiconductor materials for PEC water splitting. KEYWORDS: photoelectrochemical water splitting, TiO2 nanowires, nitrogen implantation, nitrogen doping, visible light photoactivity

T

to exhibit visible light absorption and but typically with rather limited photocatalytic activity in the visible light range.20−22 The presence of nitrogen in TiO2 has a profound effect on the photocatalysis, but the exact mechanism for the enhancement remains elusive. Indeed, there are considerable disagreements and debates within the research community regarding the exact role of N dopants due to multiple N-doping states (e.g., substitutional and interstitial).23 Additionally, most Ndoping strategies for TiO2 involve amine functional chemicals during synthesis or post thermal annealing in ammonia. These chemical doping processes are usually hard to control, because they typically involve multiple elements that could introduce unintentional dopants (e.g., by hydrogen or carbon).20−22,24,25

itanium oxide (TiO2) has been widely studied for photoelectrochemical (PEC) water splitting due to its suitable band edge positions, excellent stability, and low cost.1−9 However, the PEC performance of TiO2 to date is severely limited by its poor efficiency in the visible light range due to its large band gap.3,10−13 Impurity doping has been widely explored to modify the optical and electronic properties of TiO2 and for improving its optical absorption in the visible range.11,14−17 For example, Chen et al. and Wang et al. reported the synthesis of black TiO2 nanoparticles and nanowire (NW) arrays through a hydrogenation method and studied their photoactivity and PEC performance.15,18 Although the hydrogenation treatment is effective in modifying the optical properties of TiO2, its visible light photoactivity remains negligibly small. Cho et al. reported carbon (C) and tungsten (W) codoped TiO2 NW arrays with significantly improved photocurrent for water splitting, but with little visible light photoactivity.19 Nitrogen (N)-doped TiO2 has also been shown © 2015 American Chemical Society

Received: April 21, 2015 Revised: May 20, 2015 Published: June 8, 2015 4692

DOI: 10.1021/acs.nanolett.5b01547 Nano Lett. 2015, 15, 4692−4698

Letter

Nano Letters

Figure 1. Structure characterization of TiO2 and N-TiO2 NWs. SEM, TEM, HRTEM, and FFT images of pristine TiO2 (a, b, c, and d), nitrogen implanted TiO2 with nitrogen dose of 7 × 1014 (e, f, g, and h) and 1 × 1016 ions/cm2 (i, j, k, and l).

microscopy (TEM), and UV−vis absorption measurements were conducted before and after nitrogen implantation to probe the effects of nitrogen implantation on the morphology, structure, and optical properties of the TiO2 NWs. SEM image shows the as-prepared NW arrays are nearly vertically aligned on FTO substrates, with a rectangular cross section and a smooth surface (Figure 1a). TEM studies further confirm that the TiO2 NWs exhibit a smooth surface (Figure 1b) with wellresolved lattice fringes (Figure 1c) and perfectly symmetrical fast Fourier transform (FFT) pattern (Figure 1d), suggesting the as-prepared TiO2 are single crystals. With the increase of nitrogen-implanted dosage, the NWs exhibit increasingly rougher surface as seen in SEM images (Figure 1e and 1i), with increasing number of defects and worsening crystalline structure as seen in TEM and FFT studies (Figure 1f,g,h and j,k,l, respectively). XRD studies indicate the structure of TiO2, as implanted N-TiO2 and post-annealed N-TiO2 remains rutile (Figure S1, Supporting Information), suggesting this treatment does not change the overall structure phase of the TiO2 NWs. With nitrogen implantation, the color of the TiO2 NW arrays is changed from white to green, and further changed to brilliant yellow after annealing at 350 °C in air for 10 min (Figure 2a inset), demonstrating that the optical property of TiO2 is greatly modified by nitrogen implantation and post-thermal annealing process. We also conducted UV−vis diffuse reflectance measurement to study the optical properties of the TiO2 samples (Figure 2a). The light absorption edge of the pristine TiO2 is located around 418 nm, consistent with the band gap of rutile TiO2 (3.0 eV).15 UV−vis spectrum of the asimplanted N-TiO2 shows a rather broad absorption in the range from 400 to 700 nm, indicating a broad distribution of localized energy levels within the band gap. Importantly, the postimplantation annealed N-TiO2 shows an apparent absorption edge at around 530 nm, suggesting a well-defined energy level within the band gap after annealing. These studies clearly demonstrate that the nitrogen implantation can fundamentally

Such complicated doping makes it difficult to fundamentally probe the underlying mechanism of nitrogen dopants in TiO2. Alternatively, an ion-implantation approach has been reported for introducing N dopant in anatase TiO2 nanotubes, yet it has limited improvement in visible photoactivity (e.g., IPCE < 2% at 450 nm).26 With a polycrystalline/amorphous structure, this system is plagued with extensive defects and recombination sites that prevent a fundamental understanding of the exact role of the N-doping for further optimizing the performance of TiO2. Ion implantation is a typical material engineering process, commonly used to change the physical, chemical and electronic properties of semiconductor materials in industry.27,28 Comparing with other doping approaches, the ion implantation process does not involve any elements other than the element of interest and could offer a much cleaner and more reliable method for introducing selected impurities into a solid state material. It is also a general approach for introducing different or multiple impurity elements for systematic studies. Additionally, it is beneficial to use single crystal photoelectrodes to minimize electron/hole recombination at the polycrystalline grain boundaries during charge transport process. Compared to anatase TiO2, single crystal rutile TiO2 photoelectrodes have shown better PEC water splitting performance due to its more efficient light absorption.11,15,29,30 Here we report the nitrogen implantation doping of single crystalline rutile TiO2 nanowire arrays as photoanodes for PEC water splitting, and demonstrate the best PEC efficiency in the visible light range among all TiO2-based materials reported to date. Rutile TiO2 nanowire (NW) arrays were synthesized on fluorine-doped tin oxide (FTO) glass substrate using a previously reported hydrothermal method.15,31 Nitrogen was implanted into TiO2 NW arrays at room temperature with different ion doses, using a metal vapor vacuum arc (MEVVA) ion source implanter (see Methods in Supporting Information). Scanning electron microscopy (SEM), transmission electron 4693

DOI: 10.1021/acs.nanolett.5b01547 Nano Lett. 2015, 15, 4692−4698

Letter

Nano Letters

irreparable defects that can function as the recombination centers to degrade the overall photoactivity. To quantify the photoactivity as a function of wavelength and probe the effect of nitrogen implantation, we have conducted the incident-photon-to-current-conversion efficiency (IPCE) measurement on pristine TiO2, as-prepared N-TiO2 and postannealed N-TiO2 NWs at 0.5 V vs Ag/AgCl. The IPCE can be described using the equation IPCE = (1240 × I)/(λ × Jlight), where I is the measured photocurrent density at specific incident wavelengths, λ is the wavelength of incident light, and Jlight is the measured light intensity of the incident light. The IPCE spectra of pristine TiO2, as-prepared N-TiO2 and postannealed N-TiO2 NWs are shown in Figure 2d. The pristine TiO2 exhibits a maximum IPCE of 18% at around 380 nm with the absorption edge located at around 420 nm, consistent with literature results for rutile TiO2 NWs.15,21,30 Although the asprepared N-TiO2 shows decreased photoactivity in UV region, it exhibits slightly increased IPCE values in visible light region. Significantly, the post-annealed N-TiO2 exhibits substantially enhanced IPCE over the UV region and unprecedented visible light photoactivity beyond 420 nm. The maximum IPCE of NTiO2 at visible region is ∼10% at 450 nm, which is much higher than ∼0.2% for the pristine TiO2 NWs and 0.4% for the asprepared N-TiO2 NWs. The IPCE of 10% achieved at 450 nm in our sample represents the highest value among the all reported TiO2-based photoanodes without loading of other cocatalysts for PEC water splitting. It is considerably higher than ∼1% for previously reported hydrogen-treated TiO2 NWs,15 ∼ 2% for nitrogen-doped TiO2 NWs,20 ∼ 3.5% for hydrogen and nitrogen cotreated TiO2 NWs,21 ∼ 1.5% for W and C codoped TiO2 NWs,19 and 4% for core−shell sulfurdoped black TiO2.32 Additionally, we also demonstrated that nitrogen implantation can also improve the visible light photoactivity of other TiO2 nanostructures including anatase TiO2 NWs and commercial P25 for PEC water splitting (Figure S4, Supporting Information), suggesting the nitrogen implantation is a general method to improve the visible light photoactivity of TiO2. In order to understand the effect of the post-implantation annealing on the PEC performance of N-TiO2 NWs, we have employed X-ray photoelectron spectroscopy (XPS) to probe the surface chemical states in N-TiO2 NWs. The XPS N 1s spectra of the as-implanted N-TiO2 show several broad peaks located in the range of 395−402 eV (Figure 3a, lower panel), indicating nitrogen has multiple bonding states after implantation into TiO2. A multi-Gaussian fitting of the XPS N 1s curves (Figure 3a) reveals three peaks centered at around 396.7, 400.1, and 402.4 eV, which can be assigned to the chemical bonding of N−Ti, N−O, and N−N,33 suggesting the physical implantation process can result in multiple nitrogen dopant states and the interstitial N−O and N−N are the major dopant states. Interestingly, the XPS N 1s profiles of N-TiO2 before and after post-implantation annealing (Figure 3a, upper panel) show that the ratio of interstitial N−O and N−N states are substantially decreased from 76% to 25% of total nitrogen dopants after the thermal annealing process. In the annealed sample, the substitutional N−Ti becomes the dominant dopant state. Comparing the XPS Ti 2p spectra of the as-implanted NTiO2 and the annealed N-TiO2 with that of the pristine TiO2 show that a broadening and negative shift in binding energy after nitrogen implantation (Figure 3b, upper panel). The two main peaks located at 457.2 and 464.5 eV are the characteristics of TiO2. The fitting curves of XPS Ti 2p (Figure 3b, lower

Figure 2. Characterization and PEC performance of TiO2 and N-TiO2 NWs. (a) Light absorption properties of TiO2, as prepared N-TiO2, and post-annealed N-TiO2. The insets are the digital pictures of these samples. (b) Linear sweep voltammograms of pristine TiO2, as prepared N-TiO2 and post-annealed N-TiO2 under 100 mW/cm2 xenon light illumination with a scan rate of 20 mV/s in 1.0 M NaOH aqueous electrolyte. (c) Photoresponse of TiO2, as-prepared N-TiO2 and post-annealed N-TiO2 NWs under chopped visible light illumination (>400 nm) at 0.5 V vs Ag/AgCl. (d) IPCE spectra of TiO2, as-prepared N-TiO2 and post-annealed N-TiO2 NWs collected at 0.5 V vs Ag/AgCl.

change the electronic structure and optical properties of TiO2, and produce N-TiO2 with substantial visible light absorption. PEC measurements were carried out in a three-electrode electrochemical system (see Methods and setup image in Figure S2, Supporting Information). We have first conducted linear sweep voltammogram measurements under dark and light irradiation. These studies show that the pristine TiO2 and the as-implanted N-TiO2 without post-implantation thermal annealing exhibit very similar photocurrent (Figure 2b), indicating no obvious photocurrent enhancement in the asimplanted sample, despite their distinct optical absorption characteristics. In contrast, a significantly enhanced photocurrent is observed for the post-implantation annealed N-TiO2 (Figure 2b). The photocurrent density of the annealed N-TiO2 is ∼1.92 mA/cm2 at 0.5 V vs Ag/AgCl, which is more than four times larger than that of the pristine and the as-implanted TiO2 NWs. More importantly, the annealed N-TiO2 shows unprecedented visible light photoactivity (Figure 2c), with an overall photocurrent density of ∼0.49 mA/cm2 at 0.5 V vs Ag/ AgCl in the visible range, greatly exceeding the best value (∼0.2 mA/cm2) reported previously for nitrogen doped TiO2.20 Overall, the photocurrent density obtained from the postimplantation annealed N-TiO2 NWs under visible light illumination is more than 1 order of magnitude higher than that of pristine TiO2 NWs. We also investigated the effects of nitrogen implantation dosage and annealing temperature on PEC performance (Figure S3, Supporting Information). The maximum photocurrent density was obtained from N-TiO2 NWs with an implantation ion density of 7 × 1014/cm2, at a post-implantation annealing temperature of 350 °C. Further increase of nitrogen implantation dosage and annealing temperature can lead to a decrease in photoactivity. The excessive implantation doping could create a large number of 4694

DOI: 10.1021/acs.nanolett.5b01547 Nano Lett. 2015, 15, 4692−4698

Letter

Nano Letters

Methods in Supporting Information).37,38 We consider the 2 × 2 × 3 periodic supercells of pristine TiO2 (Figure 3c(1)), and TiO2 with multiple dopant states including interstitial N, interstitial N2, and substitutional N (Figure 3c(2−4)). The escape energies of different nitrogen dopants in TiO2 are calculated (listed in Figure 3d). The larger escape energy indicates the more stable specie. For interstitial N and N2 doped TiO2, the escape energies for N atom from N−O bond and N−N−O bond are 4.64 eV and −5.84 eV, respectively. The negative escape energy indicates that the escape of N2 is energy favorable and dominating process. However, the escape energy for the N atom from the substitutional sites is 10.06 eV, which is unfavorable to occur, compared to the escape of interstitutional N dopants. These calculations clearly suggest that interstitial N2 is rather labile and the substitutional N is the most stable state in N-TiO2. Additionally, the escaped interstitial-N species during the post-implantation annealing process may also be captured by oxygen vacancy defects in TiO2 to produce substitutional-N, which is the reverse process of substitutional-N atom escaping from TiO2 lattice and is energy favorable. As a result, the number of interstitial-N dopants may decrease while the substitutional-N dopants can increase during the thermal annealing process. Importantly, these calculations are consistent with our experiment results that the ratio of substitutional N to Ti (N/Ti) increased from 2.5/100 to 4.0/100 after the annealing process. Together, our studies clearly demonstrate that the post-implantation annealing of N implanted-TiO2 NWs could selectively enrich substitutional N-dopant concentration and lead to a substantial increase of the PEC performance in the visible range. Typically, the PEC performance is dependent on light absorption, charge separation in the photoanode materials and charge injection from the material surface to electrolyte (as illustrated in Figure S5, Supporting Information). UV−vis spectra indicated the visible light absorption was greatly improved after nitrogen implantation. To probe the fundamental roles of nitrogen doping in PEC enhancement, we further decoupled the charge separation and injection efficiency by using H2O2 as the hole scavenger. Previous studies have shown that that H2O2 can be used as a hole scavenger to

Figure 3. XPS characterization and the theoretical calculation of the nitrogen dopant chemical states. (a) Normalized and fitted XPS N 1s spectra of post-implantation annealed N-TiO2 (Upper) and asprepared N-TiO2 (Lower). (b) Upper: normalized XPS Ti 2p spectra of air annealed pristine TiO2, as-prepared N-TiO2 and post-annealed N-TiO2. Lower: the fitting curves for XPS Ti 2p of as-prepared NTiO2. (c) Relaxed structures of (1) pristine TiO2, (2) interstitial Ndoped TiO2, (3) interstitial N2-doped TiO2, (4) substitutional Ndoped TiO2. (d) Calculated escape energies of the different nitrogen chemical states.

panel) show two small new peaks centered at 457.9 and 463.4 eV, indicating the existence of Ti3+ after nitrogen implantation.34 These studies clearly demonstrate that the nitrogen implantation with post-implantation annealing can produce predominantly substitutional nitrogen states. To further understand the evolution of nitrogen species in the as-implanted and the post-implantation annealed TiO2 samples, we have carried out theoretical calculations to probe the stability of different nitrogen states and oxygen vacancies in rutile TiO2. The calculations are performed by using SIESTA package35,36 based on density-functional theory (DFT) (see

Figure 4. Charge injection, separation efficiency, electrical properties and average photoexcited carrier lifetime. (a) Charge injection (C−I) efficiency of TiO2 and N-TiO2 NWs as a function of potentials. (b) Charge separation (C−S) efficiency of TiO2 and N-TiO2 NWs as a function of potential. (c) Mott−Schottky plots of TiO2 and N-TiO2 NWs at the frequency of 10 kHz under dark condition. (d) Electrical properties of single TiO2 and NTiO2 NW. (e) Photovoltage decay curves of TiO2 and N-TiO2 NWs. (f) Average lifetime distribution of photoexcited carriers in TiO2 and N-TiO2 NWs. 4695

DOI: 10.1021/acs.nanolett.5b01547 Nano Lett. 2015, 15, 4692−4698

Letter

Nano Letters

importantly, this approach provides unique insights to the extended dynamics of long-lived photoexcited carriers, which can be used in the water splitting reaction, because this lifetime scale commensurates with the time scale required for chemical reactions.42 Overall, the average lifetime of photoexcited carriers in N-TiO2 is several times longer than that in pristine TiO2 (Figure 4f). In general, longer carrier lifetime should enhance the PEC performance in both dye-sensitized solar cells and PEC water-splitting cells43,44 because the prolonged lifetime implies the increased proability for the photoexcited carriers to be used for photochemical reactions. The significant improvement of PEC performance of N-TiO2 suggests our nitrogen implantation method is an effective method to modify the optical and electronic properties of TiO2 for PEC water splitting application. Electrochemical and photoelectrochemical studies demonstrate the enhancement is due to the improved light absorption in the visible light range and more efficient charge separation induced by the improved electrical conductivity in N-TiO2. However, the studies on charge injection efficiency indicate N-implantation does not obviously improve the surface reaction kinetics. To further improve the charge injection performance, we have thus employed cobalt hydroxide as an oxygen evolution reaction (OER) catalyst to modify the N-TiO2. The cobalt OER catalyst was deposited on TiO2 and N-TiO2 NWs using a reported dipcoating method.45 Figure 5a shows the linear sweep voltammo-

harvest photoexctied holes with nearly 100% efficiency on hematite.39 With the assumption that the hole collection efficiency is 100% when H2O2 is used as the hole scavenger, it is possible to estimate the charge injection efficiency by calculating the photocurrent ratio measured without/with H2O2. In this way, the charge separation efficiency can also be estimated by dividing photocurrent density measured with H2O2 by the light absorption efficiency (see Methods in Supporting Information). Figure 4a and b show the charge injection and separation efficiency of TiO2 and N-TiO2 NWs. We did not observe obvious enhancement of the charge injection efficiency after N-implantation, suggesting the Nimplantation does not significnatly change the surface water oxidation kinetics. On the other hand, it is imporant to note that the charge separation efficiency measured under 370 nm light illumination, where both TiO2 and N-TiO2 exhibit similar absorption behavior, has been signficantly improved in N-TiO2 NWs. Importantly, under visible light (450 nm), the N-TiO2 NWs also show respectable charge separation efficiency (∼20− 50%), whereas the pristine TiO2 NWs does not show obvious visible light repsonse (Figure S6, Supporting Information). To understand how the N-implantation improves the charge separation efficiency, we have conducted electrochemical impedance studies. The Mott−Schottky plots of both TiO2 and N-TiO2 samples show postive slopes, consistent with the expected n-type semiconductor characteristics. Importantly, the N-TiO2 NWs show subtantially smaller slopes, compared to the pristine TiO2 NW samples, suggesting an increase of donor densities in TiO2 NWs after nitrogen implantation. The donor density can be calculated from the Mott−Schottky plots using the equation Nd = (2/e0εε0)[d(1/C2)/dV]−1, where e0 is the electron charge, ε the dielectric constant of TiO2 (ε = 170), ε0 the permittivity of vacuum, Nd the donor density, and V the potential at the electrode.40 On the basis of these analyses, the N-implantation process leads to an increase of the donor density by about 1 order of magnitude (2.05 × 1019 for pristine TiO2 NWs and 1.54 × 1020 cm−3 for N-TiO2 NWs). To further confirm N doping changes the electrical properties, we have also measured the electrial conductance of individual TiO2 NWs before and after N-implantation (Figure 4d). Importantly, the electrical conductivity of N-TiO2 NWs (176 S/m) is about 1 order of magnitude higher than that of printine TiO2 NWs (16 S/m), consistent with the Mott−Schottky studies. The increased donor density may be attributed to the creation of Ti3+ state, a shallow donor for TiO2, during N-implantation process, which has been demonstrated in the XPS studies. Additionally, the increased carrier density can also increase the band bending at the surface of TiO2 and, thus, improve the charge separation in TiO2.15,40,41 To further understand the PEC perfomance, we investigated the electron recombination kinetics of TiO2 NWs and N-TiO2 NWs by monitoring the photovoltage decay as a funtion of time upon turning off the illumination (Figure 4e). It is evident that the N-TiO2 has a slower photovoltage decay rate, compared to prisitne TiO2. The photovoltage decay curves can be used to derive the average lifetime of the photogenerated carrier (τn), accroding to the equation: τn = −((κBT)/(e))((dV)/(dt))−1, where κB is the Boltzmann constant and T is the temperature. Compared to ultrafast spectroscopy in femto- or nanosecond scale, this measurement preferentially probes the slow recombination kinetics (milliseconds to seconds) of interfacial charge transfer processes occurring at the semiconductor−liquid junctions. More

Figure 5. PEC performance of OER/N-TiO2 NWs. (a) Linear sweep voltammograms of N-TiO2 and OER/N-TiO2 NWs under dark and 100 mW/cm2 xenon light illumination with a scan rate of 20 mV/s. (b) Comparison of charge injection (C−I) efficiency of N-TiO2 and OER/N-TiO2 NWs. (c) IPCE spectra of TiO2, OER/TiO2, N-TiO2 and OER/N-TiO2 NWs collected at 0.5 V vs Ag/AgCl. (d) Simulated photocurrent densities for the TiO2, OER/TiO2, N-TiO2, and OER/ N-TiO2, respectively, by integrating their IPCE spectra with a standard AM 1.5G solar spectrum (ASTM G-173-03).

grams of N-TiO2 NWs with/without OER catalyst under dark and light illumination. Under dark conditions, the linear sweep voltammograms of OER modified N-TiO2 NW show lower current onset potential, indicating OER modification can effectively reduce the overpotential and improve the surface water oxidation kinetics. Importantly, the photocurrent density of OER modified N-TiO2 NWs is significantly increased in the whole studied potential range. The photocurrent density of OER/N-TiO2 at 0.5 V vs Ag/AgCl can reach ∼2.45 mA/cm2, 4696

DOI: 10.1021/acs.nanolett.5b01547 Nano Lett. 2015, 15, 4692−4698

Letter

Nano Letters in comparison to 1.9 mA/cm2 for N-TiO2. Figure S7 (Supporting Information) shows the linear sweep voltammograms of N-TiO2 and OER/N-TiO2 under visible light illumination. A photocurrent density of 0.83 mA/cm2 can be achieved at 0.5 V vs Ag/AgCl. To prove our hypothesis that OER catalyst can improve the charge injection efficiency, we further compared the charge injection efficiency of N-TiO2 before and after OER catalyst modification (Figure 5b). It is clear that the charge injection has been substantially improved with the assistance of OER catalyst and it is almost 100% when the potential is larger than 0 V vs Ag/AgCl. The average lifetime of photoexcited carriers in OER/N-TiO2 is also longer than that in N-TiO2 (Figure S8, Supporting Information), suggesting OER catalyst can also increase the proability for the photoexcited carriers to be used for photochemical reactions. We also measured the IPCE spectra of TiO2 and N-TiO2 with/without OER catalyst at 0.5 V vs Ag/AgCl (Figure 5c). The IPCE values for both TiO2 and N-TiO2 have been increased with the assistance of OER catalyst. Significantly, the IPCE at 450 nm can reach a highest value of 17% for OER/N-TiO2 NWs, which is much higher than 10% obtained in N-TiO2 NWs and also among the best values reported for the TiO2 based photoelectrodes to date (Table S1, Supporting Information).19−21,29,46 Considering the irradiance spectrum may vary for different light sources (e.g., lamp lifetime, power, and filters) used in the measurements; it is difficult to compare the overall photocurrent densities reported by different research laboratories. In this regard, IPCE is independent from light sources and can be used to calculate the photocurrent density under standard solar light by integrating the IPCE with a standard AM 1.5G solar spectrum (ASTM G-173-03), using the equation:15 600

I=

∫300

The optimized nitrogen implanted TiO2 NW arrays without surface cocatalyst yield an IPCE of 10% at 450 nm, which is the best results among all TiO2 based photoelectrodes reported to date. The enhanced PEC performance is largely attributed to the increased visible light absorption and improved charge separation efficiency. With the assistance of OER catalyst and improvement of the charge injection efficiency, a further increased IPCE of 17% can be achieved at 450 nm. The capability of making TiO2 photoelectrodes with significant visible light photoactivity using ion implantation could open up exciting new opportunities in various areas, including PEC water splitting, dye sensitized solar cells, and photocatalysis.



ASSOCIATED CONTENT

S Supporting Information *

Methods, SEM images, XRD, and PEC data are shown. The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.nanolett.5b01547.



AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected]. *E-mail: [email protected]. Author Contributions ¶

These authors contributed equally to this work.

Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS We acknowledge the support from the U.S. Department of Energy, Office of Basic Energy Sciences, Division of Materials Science and Engineering through Award DE-SC0008055. We acknowledge Electron Imaging Center for Nanomachines (EICN) at UCLA for the support of TEM, supported with funding from NIH-NCRR Shared Resources Grant (CJX1443835-WS-29646) and NSF Major Research Instrumentation Grant (CHE-0722519). X.H.X. also acknowledges the support from NSFC (U1260102, 51371131) and NCET (12-0418).

1 λ IPCE(λ)E(λ)d(λ) 1240

Figure 5d shows the integrated photocurrent density as a function of wavelength in the range from 300 to 600 nm. The TiO2, OER/TiO2, N-TiO2, and OER/N-TiO2 have achieved a photocurrent density of 0.245 mA/cm2, 0.41 mA/cm2, 1.4 mA/ cm2 and 1.9 mA/cm2, respectively. Between 420 and 600 nm, there are substantial increases in overall photocurrent densities for both N-TiO2 and OER/N-TiO2, but almost no increase for the TiO2 and OER-TiO2 in the integration curves, demonstrating the unprecedented visible light contribution to the PEC performance in N-TiO2 (Figure 5d). Overall, the photocurrent density of 1.9 mA/cm2 represents the best performance achieved on TiO2 photoelectrodes ever. Furthermore, we have also studied the long-term stability of the visible light photoactivity of the N-TiO2 NWs (Figure S9, Supporting Information). In comparison with most of previous studies about N-doped TiO2 that did not report long-term stability,20,21 the OER/N-TiO2 can maintain more than 70% of initial photocurrent after continuous illumination for 7 days, suggesting its excellent photostability for water splitting. On the contrast, bare N-TiO2 without OER catalyst can only maintain less than 40% of initial photocurrent after a week. In summary, we have demonstrated that nitrogen implantation can greatly improve the PEC performance of TiO2 NW photoanodes and achieve unprecedented visible light photoactivity. Our studies indicate that post-implantation thermal annealing process is critical for PEC performance and XPS studies indicate the substitutional nitrogen dopants become the dominant species after post-implantation thermal annealing.



REFERENCES

(1) Fujishima, A.; Honda, K. Nature 1972, 238 (5358), 37−38. (2) Khan, S. U. M.; Al-Shahry, M.; Ingler, W. B. Science 2002, 297 (5590), 2243−2245. (3) Chen, X. B.; Mao, S. S. Chem. Rev. 2007, 107 (7), 2891−2959. (4) Qu, Y. Q.; Duan, X. F. Chem. Soc. Rev. 2013, 42 (7), 2568−2580. (5) Gratzel, M. Nature 2001, 414 (6861), 338−344. (6) Walter, M. G.; Warren, E. L.; McKone, J. R.; Boettcher, S. W.; Mi, Q. X.; Santori, E. A.; Lewis, N. S. Chem. Rev. 2010, 110 (11), 6446− 6473. (7) Schneider, J.; Matsuoka, M.; Takeuchi, M.; Zhang, J. L.; Horiuchi, Y.; Anpo, M.; Bahnemann, D. W. Chem. Rev. 2014, 114 (19), 9919− 9986. (8) Wang, X. D.; Li, Z. D.; Shi, J.; Yu, Y. H. Chem. Rev. 2014, 114 (19), 9346−9384. (9) Wang, Y. C.; Tang, J.; Peng, Z.; Wang, Y. H.; Jia, D. S.; Kong, B.; Elzatahry, A. A.; Zhao, D. Y.; Zheng, G. F. Nano Lett. 2014, 14 (6), 3668−3673. (10) Wang, G. M.; Ling, Y. C.; Wang, H. Y.; Lu, X. H.; Li, Y. J. Photochem. Photobiol., C 2014, 19, 35−51. (11) Xu, M.; Da, P. M.; Wu, H. Y.; Zhao, D. Y.; Zheng, G. F. Nano Lett. 2012, 12 (3), 1503−1508. (12) Pu, Y. C.; Wang, G. M.; Chang, K. D.; Ling, Y. C.; Lin, Y. K.; Fitzmorris, B. C.; Liu, C. M.; Lu, X. H.; Tong, Y. X.; Zhang, J. Z.; Hsu, Y. J.; Li, Y. Nano Lett. 2013, 13 (8), 3817−3823. 4697

DOI: 10.1021/acs.nanolett.5b01547 Nano Lett. 2015, 15, 4692−4698

Letter

Nano Letters (13) Zhang, Z. H.; Zhang, L. B.; Hedhili, M. N.; Zhang, H. N.; Wang, P. Nano Lett. 2013, 13 (1), 14−20. (14) Liu, C.; Dasgupta, N. P.; Yang, P. D. Chem. Mater. 2014, 26 (1), 415−422. (15) Wang, G. M.; Wang, H. Y.; Ling, Y. C.; Tang, Y. C.; Yang, X. Y.; Fitzmorris, R. C.; Wang, C. C.; Zhang, J. Z.; Li, Y. Nano Lett. 2011, 11 (7), 3026−3033. (16) Liu, B.; Chen, H. M.; Liu, C.; Andrews, S. C.; Hahn, C.; Yang, P. D. J. Am. Chem. Soc. 2013, 135 (27), 9995−9998. (17) Wang, Y. C.; Zhang, Y. Y.; Tang, J.; Wu, H. Y.; Xu, M.; Peng, Z.; Gong, X. G.; Zheng, G. F. ACS Nano 2013, 7 (10), 9375−9383. (18) Chen, X. B.; Liu, L.; Yu, P. Y.; Mao, S. S. Science 2011, 331 (6018), 746−750. (19) Cho, I. S.; Lee, C. H.; Feng, Y. Z.; Logar, M.; Rao, P. M.; Cai, L. L.; Kim, D. R.; Sinclair, R.; Zheng, X. L. Nat. Commun. 2013, 4, 8. (20) Hoang, S.; Guo, S. W.; Hahn, N. T.; Bard, A. J.; Mullins, C. B. Nano Lett. 2012, 12 (1), 26−32. (21) Hoang, S.; Berglund, S. P.; Hahn, N. T.; Bard, A. J.; Mullins, C. B. J. Am. Chem. Soc. 2012, 134 (8), 3659−3662. (22) Asahi, R.; Morikawa, T.; Ohwaki, T.; Aoki, K.; Taga, Y. Science 2001, 293 (5528), 269−271. (23) Dunnill, C. W.; Parkin, I. P. Dalton Trans. 2011, 40 (8), 1635− 1640. (24) Yang, X. Y.; Wolcott, A.; Wang, G. M.; Sobo, A.; Fitzmorris, R. C.; Qian, F.; Zhang, J. Z.; Li, Y. Nano Lett. 2009, 9 (6), 2331−2336. (25) Cong, Y.; Zhang, J. L.; Chen, F.; Anpo, M. J. Phys. Chem. C 2007, 111 (19), 6976−6982. (26) Ghicov, A.; Macak, J. M.; Tsuchiya, H.; Kunze, J.; Haeublein, V.; Frey, L.; Schmuki, P. Nano Lett. 2006, 6 (5), 1080−1082. (27) Delecroix, P. Electron. Microelectron. Ind. 1971, 141, 31−37. (28) Di, Z. F.; Wang, Y. Q.; Nastasi, M.; Shao, L.; Lee, J. K.; Theodore, N. D. Appl. Phys. Lett. 2008, 93 (10), 3. (29) Yang, C. Y.; Wang, Z.; Lin, T. Q.; Yin, H.; Lu, X. J.; Wan, D. Y.; Xu, T.; Zheng, C.; Lin, J. H.; Huang, F. Q.; Xie, X. M.; Jiangl, M. H. J. Am. Chem. Soc. 2013, 135 (47), 17831−17838. (30) Hwang, Y. J.; Hahn, C.; Liu, B.; Yang, P. D. ACS Nano 2012, 6 (6), 5060−5069. (31) Liu, B.; Aydil, E. S. J. Am. Chem. Soc. 2009, 131 (11), 3985− 3990. (32) Liu, C.; Tang, J. Y.; Chen, H. M.; Liu, B.; Yang, P. D. Nano Lett. 2013, 13 (6), 2989−2992. (33) Li, J. L.; Ma, X. X.; Sun, M. R.; Li, X. M.; Song, Z. L. Thin Solid Films 2010, 519 (1), 101−105. (34) Lu, X. H.; Wang, G. M.; Zhai, T.; Yu, M. H.; Gan, J. Y.; Tong, Y. X.; Li, Y. Nano Lett. 2012, 12 (3), 1690−1696. (35) Roman-Perez, G.; Soler, J. M. Phys. Rev. Lett. 2009, 103 (9), 4. (36) Soler, J. M.; Artacho, E.; Gale, J. D.; Garcia, A.; Junquera, J.; Ordejon, P.; Sanchez-Portal, D. J. Phys.: Condens. Matter 2002, 14 (11), 2745−2779. (37) Hohenberg, P.; Kohn, W. Phys. Rev. B 1964, 136 (3B), B864− 871. (38) Kohn, W.; Sham, L. J. Phys. Rev. 1965, 140 (4A), 1133−1138. (39) Dotan, H.; Sivula, K.; Gratzel, M.; Rothschild, A.; Warren, S. C. Energy Environ. Sci. 2011, 4 (3), 958−964. (40) Wang, G. M.; Ling, Y. C.; Wang, H. Y.; Yang, X. Y.; Wang, C. C.; Zhang, J. Z.; Li, Y. Energy Environ. Sci. 2012, 5 (3), 6180−6187. (41) Pesci, F. M.; Wang, G. M.; Klug, D. R.; Li, Y.; Cowan, A. J. J. Phys. Chem. C 2013, 117 (48), 25837−25844. (42) Cowan, A. J.; Durrant, J. R. Chem. Soc. Rev. 2013, 42 (6), 2281− 2293. (43) DuChene, J. S.; Sweeny, B. C.; C, J.-P. A.; Su, D.; Stach, E. A.; Wei, W. Angew. Chem., Int. Ed. 2014, 53 (30), 7887−7891. (44) Wang, M. K.; Bai, J.; Le Formal, F.; Moon, S. J.; Cevey-Ha, L.; Humphry-Baker, R.; Gratzel, C.; Zakeeruddin, S. M.; Gratzel, M. J. Phys. Chem. C 2012, 116 (5), 3266−3273. (45) Kay, A.; Cesar, I.; Gratzel, M. J. Am. Chem. Soc. 2006, 128 (49), 15714−15721. (46) Park, J. H.; Kim, S.; Bard, A. J. Nano Lett. 2006, 6 (1), 24−28.

4698

DOI: 10.1021/acs.nanolett.5b01547 Nano Lett. 2015, 15, 4692−4698