Oriented Growth of Sc-Doped Ta3N5 Nanorod Photoanode Achieving

Subscriber access provided by University of South Dakota

Article 3

5

Oriented Growth of Sc-Doped TaN Nanorod Photoanode Achieving Low-Onset-Potential for Photoelectrochemical Water Oxidation Lang Pei, Bihu Lv, Shuangbao Wang, Zhentao Yu, Shicheng Yan, Ryu Abe, and Zhigang Zou ACS Appl. Energy Mater., Just Accepted Manuscript • DOI: 10.1021/acsaem.8b00809 • Publication Date (Web): 30 Jul 2018 Downloaded from http://pubs.acs.org on August 2, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 27 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Energy Materials

Oriented Growth of Sc-doped Ta3N5 Nanorod Photoanode Achieving Low-Onset-Potential for Photoelectrochemical Water Oxidation

Lang Pei,† Bihu Lv,‡ Shuangbao Wang,§ Zhentao Yu,† Shicheng Yan*,† Ryu Abe,∥ Zhigang Zou†,‡ †

Eco-materials and Renewable Energy Research Center (ERERC), Collaborative Innovation Center of Advanced Microstructures, College of Engineering and Applied Sciences, Nanjing University, No. 22 Hankou Road, Nanjing, Jiangsu 210093, P. R. China ‡ Jiangsu Key Laboratory for Nano Technology, National Laboratory of Solid State Microstructures, Department of Physics, Nanjing University, No. 22 Hankou Road, Nanjing, Jiangsu 210093, P. R. China. § Collaborative Innovation Center of Sustainable Energy Materials, Guangxi University, Nanning, Guangxi 530004, P. R. China ∥ Department of Energy and Hydrocarbon Chemistry, Graduate School of Engineering, Kyoto University, Katsura, Nishikyo-ku, Kyoto 615-8510, Japan

Keywords: photoelectrochemical, tantalum nitride, doping, onset potential, water splitting Abstract: Developing photoanodes that can split water with low (or even without) externally-applied bias is a critical challenge for achieving efficient solar driven photoelectrochemical water oxidation. Here, we proposed a flux-assisted oriented crystal growth route to minimize the drawbacks of Ta3N5 photoanode, including: (1) Crystallographic-oriented growth minimizes the negative effect of electronic structure anisotropy of the Ta3N5, facilitating the directional fast charge transfer in having lighter carriers effective mass of (b,c)-axes directions; (2) reducing crystal defect suppressed surface recombination of carriers and released the surface Fermi level pinning effect; (3) increment of oxygen-impurity electron donor by Sc doping

ACS Paragon Plus Environment

ACS Applied Energy Materials 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

improved the electrical conductivity, promoting the bulk charge separation and transfer. Specifically, the onset potential of Ta3N5 photoanode was dramatically negatively shifted to 0.4 VRHE, and thereby affording a 0.82% half-cell solar-to-hydrogen efficiency. 1. Introduction Photoelectrochemical (PEC) water splitting is a potential way to producing clean and sustainable hydrogen fuels.1,2 Solar driven water oxidation half-reaction is regarded to be the key step for overall water splitting, primarily due to the high overpotential through four-electron extraction from water molecules.3 Thus, the photoanode material used in practically large-scale PEC water splitting should have a moderate band gap for harvesting wide range of sunlight, a suitable band levels to oxidize water with low (or even without) external applied bias, and a capability of highly efficient charge separation and transfer.4 Tantalum nitride (Ta3N5) is proposed as one of the outstanding candidate of such photoanode materials, which can the kinetically sluggish water oxidation half-reaction due to its suitable band level, a theoretical maximum solar-to-hydrogen (STH) conversion efficiency over 15%.5 However, one of the major limitation of Ta3N5 photoanode is its relatively positive water oxidation onset potential as high as 0.8-1.0 VRHE,6,7 which is 1.0-1.2 V far from the flat band potential of Ta3N5 (-0.2 VRHE).8 Such positive onset potential the application of highly-positive potentials (or external bias in the case of practical two-electrode systems) to realize the desired photocurrent density (10 mA cm-2), lowering the STH conversion efficiency thus far. Recently, after modifying Ta3N5


Page 2 of 27

Page 3 of 27 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


photoanode by using Ni(OH)x/ferrhydrite hole-storage layers to promote interfacial charge transfer, a TiOx blocking layer to reduce surface electron-hole recombination and (Co, Ir)-containing molecular catalysts to increase charge injection efficiency, the photocurrent onset potential was still as high as 0.65 VRHE.9 Doping of Ba2+ species or co-doping by Mg and Zr into the lattice of Ta3N5 also can move the photocurrent potential reaching to about 0.6 VRHE with the help of a water oxidation cocatalyst, due to that the doping ions enhanced the electrical conductivity and improved the surface properties.5,10 However, the reasons that induce high photocurrent onset potential of Ta3N5 material still keep unclear, limiting the development of efficient Ta3N5-based photoanodes. The origin of such positive onset potentials of Ta3N5-based photoanodes generally involves a mix of factors such as the inefficient carrier separation and transportation, surface Fermi level pinning effect and the low charge injection efficiency at the interface between semiconductor photocatalyst and electrolyte.11-13 The conventional Ta3N5 particles that were prepared by high-temperature nitriding reaction include considerable amounts of crystal defects such as nitrogen (N) vacancies and reduced species. The presence of such defects have been demonstrated to accelerate the recombination between photogenerated carriers, thereby leading to the highly positive onset potential for photocurrent generation.14,15 We have indeed observed various carrier recombination processes in such Ta3N5 samples, via laser-induced photoluminescence

technique,

including

band-edge

and

deep-level

bulk

and surface trapping level recombination from surface defects.16 In addition, the



theoretical calculations have revealed that Ta3N5 has strong electric structure anisotropy, which induces a large difference in carriers effective mass between a-axis (1.35m0 for electron and 1.1m0 for hole) and (b or c)-axis (0.65m0 for electron and 0.85m0 for hole) crystallographic directions.17,18 Thus, we can conclude that reducing crystal defects and decreasing negative effect of electric structure anisotropy are potential strategies to overcome the drawbacks of Ta3N5 material and achieve highly efficient water oxidation on Ta3N5-based photoanodes. In the present study, as shown in Figure 1, we thus proposed to synthesize a-axis oriented single-crystalline nanorods of Sc-doped Ta3N5 with high crystallinity by flux-assisted nitridation. In theory, replacing Ta5+ by Sc3+ can stabilize the N3- by improving the covalency of Ta-N bond due to the lower electronegativity for Sc3+ (1.415) than Ta5+ (1.925),19 thus reducing N vacancy defects. As a result, our results demonstrated that the onset potential on the Sc-doped Ta3N5 nanorods photoanode was unprecedentedly negatively shifted up to 0.4 VRHE, which is about 400 mV lower than the conventional Ta3N5 particles photoanode, when modified the photoanode surface with Co(OH)x as a cocatalyst. The prominent photocatalytic performance is attributed to the crystallographic-oriented growth minimizing the negative effect of electronic structure anisotropy of the Ta3N5, and Sc doping promoting the significant reduction of N vacancy defects and the increment of oxygen-impurity electron donor for effective charge separation and transfer. Our oriented crystal growth design maybe offer a new concept in crystal engineering for efficient solar energy conversion.


Page 4 of 27

Page 5 of 27 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 1. A scheme to describe the oriented crystal growth for minimizing the drawbacks of Ta3N5 photocatalyst. The a-axis preferred growth single crystal with element doping can overcome the electric structure anisotropy, reduce defects and improve the electrical conductivity.

2. Experimental Preparation of photocatalysts. Sc-doped Ta3N5 nanorods were prepared by Na2CO3 flux-assisted nitriding growth method using the ScCl3 and Ta2O5 as precursors with atom ratio of Sc:Ta = 0.03, 0.05 and 0.1 (denoted as x%Sc-Ta3N5, x = 3, 5 and 10). In a typical synthesis procedure, the stoichiometry of ScCl3·6H2O (Aladdin, purity 99.9%) and Ta2O5 (Aladdin, purity 99.9%) powders were added into certain amount of ethanol. The resulting slurry was ground for 30 min and then dried in ambient air. Subsequently, the obtained powder was annealed at 650 oC for 2 h under air in a muffle furnace. After cooling to room temperature, the as-prepared oxide product was manually blended with Na2CO3 (Aladdin, purity 99.99%) with a molar ratio of 1:1 (Ta2O5:Na2CO3) in an agate mortar. After that, the precursor was carefully put in an alumina boat, heated to 950 oC with a ramp rate of 10 oC min-1 and



maintained at this temperature for 20 h in flowing NH3 (1L min–1) in a tube furnace. The obtained Sc-doped Ta3N5 powders were rinsed with distilled water, then dried at 80 oC. For comparison, using the above-mentioned nitriding process, Ta3N5 nanorods were synthesized in absence of ScCl3 and Ta3N5 particles were synthesized by directly nitriding Ta2O5. Fabrication of photoelectrodes. The Ta3N5 electrode was prepared by electrophoretic deposition (EPD) on F-doped SnO2 (FTO) glass following a previously reported procedure.6 40 mg of Ta3N5 and 10 mg of iodine were dispersed in 50 mL of acetone. The suspension was sonicated for 20 min. The EPD procedure for Ta3N5 nanorods and Sc-doped Ta3N5 nanorods was conducted between two FTO electrodes which were parallel immersed into the solution with a distance of 10 mm, and a 15 V of voltage was applied for 3 min. The film on FTO had a average thickness of about 3 µm. To produce a similar thickness of Ta3N5 particle photoanodes, a slightly higher bias (20 V) was used. The coated area of the Ta3N5 films was ca. 1 cm × 1 cm. After that, 10 µL of TaCl5 methanol solution (10 mM) was dropped onto the electrode, followed by drying at 160 oC for 3 min in air. After repeating this procedure for five times, the resulting electrodes were calcinated in flowing NH3 (0.5 L/min) at 470 oC for 30 min. To accelerate the surface reaction kinetics, the loading of Co(OH)x cocatalysts on Ta3N5 electrodes was referred to the literature.6 Typically, the colloidal Co(OH)x solution was obtained by mixing 180 µL of NaOH (1 M) and 30 mL of Co(NO3)2 (0.01 M) aqueous solution. After that, the Ta3N5 electrodes were immersed into the


Page 6 of 27

Page 7 of 27 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Co(OH)x colloidal solution for 1 h. Finally, the Ta3N5 electrodes were rinsed with deionized water, then dried at 60 oC in air. 3. Results and discussion Structure and Morphological Characterizations. Sc-doped Ta3N5 nanorods were obtained by Na2CO3 flux-assisted nitriding growth method using the ScCl3 and Ta2O5 as precursors with atom ratio of Sc:Ta = 0.03, 0.05 and 0.1 (denoted as x% Sc-Ta3N5, x = 3, 5 and 10). For comparison, using the above-mentioned nitriding process, Ta3N5 nanorods were synthesized in absence of ScCl3 and Ta3N5 particles were obtained by directly nitriding Ta2O5. As displayed in Figure 2a, comparing with the standard X-ray powder diffraction (XRD) data (JCPDS Card No. 79-1533), these as-prepared materials were assigned to the orthorhombic-phase Ta3N5. Obviously, the XRD peak position of (200) facet gradually shifted to lower 2θ value with the increase of Sc doping amount (Figure 2b), indicating the slight lattice expansion, which is mainly attributed to the partially replacing Ta5+ (64 pm) with larger ionic radii of Sc3+ (75 pm). After Sc doping, the X-ray photoelectron spectroscopy (XPS) analysis showed that the signal at 397-407 eV can be deconvoluted into two peaks (Figure 2c): 403.6 eV for Ta4p12 and 400.7 eV for Sc 2p.20,21 And N1s peaks at 395.6 and 394.2 eV were assigned to the N-Ta and N-Sc bonds, respectively. Obviously, the binding energy of N-Sc is about 1.4 eV lower than that of N-Ta, resulting from that the electronegativity of Sc3+ (1.415) is lower than that of Ta5+ (1.925).19 In addition, compared to the pure Ta3N5, the Sc doping induced 0.7 eV decrease in binding energy of N1s, further indicating that the formation of N-Sc bonds makes N atom tend to gain



electrons more easily from Sc due to its smaller electronegativity. According to previously experimental results,22,23 surface defects, such as oxygen-related defects, are inevitably formation during the high-temperature nitriding reaction of Ta3N5. Indeed, O1s core-level XPS confirmed that the surface lattice O/N of 0.74:1 for Ta3N5 increased up to 0.83:1 for 5% Sc-Ta3N5 (Figure 2d). This indicates that replacing Ta5+ by Sc3+ will introduce the more oxygen impurities into the surface lattice of Ta3N5, this is also evident in the theoretical calculation of defect formation energies. As displayed in Figure S1, we investigated the formation energies of following defects: (a) the N was substituted by O (denoted as ON); (b) the Ta was substituted by Sc (denoted as ScTa); (c) the codoping of ScTa and ON occurs (denoted as ScTa+ nON, n is the number of ON). The results indicated that both the ON-doped Ta3N5 and ScTa+2ON-codoped Ta3N5 are thermodynamically stable, agreeing with our experiment results and previous calculations.24 Moreover, an about 1:3 atomic concentration ratio of O/N (Table S2) was estimated by energy dispersive spectrometer (EDS).


Page 8 of 27

Page 9 of 27 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 2. (a) XRD and (b) enlarged XRD patterns for Ta3N5 particles and Sc-doped Ta3N5 nanorods photocatalysts. (c) N 1s, Sc 2p and Ta 4p and (d) O1s XPS spectra of 5% Sc-Ta3N5 and Ta3N5 nanorods.

Next, scanning electron microscopy (SEM) and transmission electron microscope (TEM) observations confirmed that the well-dispersed Ta3N5 and Sc-doped Ta3N5 nanorods with an average diameter of ~50 nm and a length of ~300-500 nm were successfully synthesized in the presence of Na2CO3 flux (Figure 3a, b and Figure S2). The Ta3N5 obtained by directly nitriding the Ta2O5 without flux addition presented the porous sintered particle, in good agreement with the previous report.6 The high-resolution TEM (HRTEM) image of Ta3N5 nanorod (Figure 3c) clearly indicates that, there are two interplanar d spacings, 0.516 and 0.194 nm,



respectively corresponds to the (020) and (200) lattice planes of an orthorhombic Ta3N5 phase. The selected area electron diffraction (SAED) pattern reveals that the as-grown Ta3N5 is a high quality single crystal nanorod with the longitudinal direction along [100] and the cross-section direction along [010] (Figure 3c), confirming that the Ta3N5 nanorod grew along the a-axis direction. The atomic-resolution high-angle annular dark field (HAADF) image exhibits the distribution of heaviest Ta and Sc elements (Figure 3d), further indicative of the high single-crystalline quality. Although the Sc and Ta atomic numbers are clearly different, the slight Sc concentration and randomly substituting for Ta restrict the direct observation of Sc from the HAADF image. Elemental mapping for a single Sc-doped Ta3N5 nanorod revealed that Ta, N and Sc were homogeneously incorporated into the whole Ta3N5 nanorod (Figure 3e-h). Compared to the raw feeding, slight decrease in nominal Sc content for all Sc-doped Ta3N5 nanorods was observed by EDS and XPS, probably due to the volatilization of Sc during the nitriding (Table S2).


Page 10 of 27

Page 11 of 27 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 3. Morphology and composition of 5% Sc-Ta3N5 nanorods. (a) SEM image. (b) TEM image. (c) Enlarged high-resolution TEM of 5% Sc-Ta3N5 observed on the red circle region shown in Figure 3b. Inset shows corresponding SAED pattern. (d) Atomic-resolution HAADF image of (001) surface. (e-h) Elemental mapping of N, Sc and Ta for a single nanorod.

Evaluation of PEC water oxidation performances. The Ta3N5 photoanodes were fabricated by depositing Ta3N5 powders on fluorine-doped tin oxide (FTO) substrate by electrophoretic deposition method. The PEC water splitting activity was firstly evaluated in 1 M NaOH electrolyte (pH 13.6) under 500 W Xe lamp irradiation. Obviously, the Ta3N5 nanorods exhibited the higher PEC activity than Ta3N5 particles, and the Sc doping lower than 10% further enhanced the PEC activity of Ta3N5 nanorods (Figure S3). 5% Sc-Ta3N5 exhibited the lowest photocurrent onset potential and highest photocurrent density at 1.23 VRHE. Therefore, in this study, the 5%



Sc-Ta3N5 nanorods, Ta3N5 nanorods and Ta3N5 particles were selected as model photoanodes for disclosing the nature of high photocurrent onset potential of Ta3N5.

Figure 4. (a) Current-potential curves tested in 1 M NaOH electrolyte (pH 13.6) under chopped AM 1.5G simulated sunlight. (b) UV-Vis absorption spectra and (c) the corresponding Tauc plots. (d) Photoluminescence spectra obtained by excitation wavelength of 405 nm.

Figure. 4a provides the typical current-potential (J-V) curves of the Ta3N5 photoanodes under chopped air mass (AM) 1.5G-simulated sunlight (100 mW cm-2). The photocurrent onset potential (defined as the potential where photoanodic current reaches 0.2 mA cm-2 according to refs. 9 and 25) of 5% Sc-Ta3N5 is observed at 0.7 VRHE, is 100 and 220 mV cathodic shift than those of Ta3N5 nanorods and Ta3N5 particles electrodes, respectively. In order to confirm the onset potential position, we also provided the data of J-V curves at scan rate 10 mV/s without light chopping for


Page 12 of 27

Page 13 of 27 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


all the three photoanodes, as shown in Figure S4. It was found that the steady-state photocurrent densities were equal to the photocurrent densities of the current-potential curves in Figure 4a, thus confirming that the onset potential defined from Figure 4a can represent steady-state behaviour. To understand the nature of the remarkably negative shift of photocurrent onset potential of 5% Sc-Ta3N5 and Ta3N5 nanorods, the optical absorption properties of the three samples were examined firstly. As seen in the UV/Vis absorption spectra in Figure 4b, Ta3N5 nanorods, 5% Sc-Ta3N5 nanorods and Ta3N5 particles demonstrated the little difference in absorption edge located at 576-587 nm, which was attributed to the band-band transition from the orbitals of N 2p to Ta 5d .17 Indeed, the density functional theory (DFT) calculations confirmed that Sc doping has negligible contribution to the density of state of both the bottom of conduction and top of value band (Figure S5). Tauc plot of (αhν)2 versus

hν showed that the Ta3N5 particles exhibit a band gap of 2.11 eV (Figure 4c). About 0.04 eV difference in band gap for 5% Sc-Ta3N5 than Ta3N5 particles would originate from its relatively high oxygen content in bulk, leading to downshift of the valence band maximum from the orbital hybridization of lower energy O 2p and N 2p.26 Clearly, the significant cathodic shift in onset potential was not attributable to the difference in light absorption. The photoluminescence (PL) spectrum (Figure 4d) of 5% Sc-Ta3N5 clearly shows an optical emission peak at around 592 nm, close to its absorption edge of 576 nm. Accordingly, the emission peak can be attributed to the band-band recombination of 5% Sc-Ta3N5. Similar response at 598 nm was observed for Ta3N5 nanorods with 582 nm absorption edge. However, the Ta3N5 particles with



587 nm absorption edge exhibited a quenched characteristic PL peak between 580 and 850 nm centred at 700 nm, which was attributed to a strong nonradiative recombination at the N vacancies and reduced Ta species.16 Indeed, UV-Vis spectrum of Ta3N5 particles exhibited an obvious defect-related light harvesting band range from 600 to 800 nm (Figure 4b), which has been assigned to nitrogen vacancies (VN) as well as reduced Ta species.27,28 In contrast, flux-assisted nitridation route to Ta3N5 nanorods largely suppress nitrogen vacancies, as evidenced by the flattened absorption range from 600 to 800 nm. Thus, we can conclude that the lower photocurrent onset potential on Ta3N5 nanorods can be attributed at least in part to the suppressed surface recombination on the nanorods.

Figure 5. (a-c) Mott-Schottky plots (d) Open circuit potential measurements of 5% Sc-Ta3N5, Ta3N5 nanorods and Ta3N5 particles photoanode under AM 1.5G simulated sunlight and dark conditions. Owing to that only the space charge layer capacitance was probed in Mott-Schottky (MS) measurements (Figure 5a-c), the flat band potential (Vfb) of three Ta3N5 photoanodes were estimated based on the MS equation of 1 2 kT = (Vapp − Vfb - ) Csc2 A2εε 0 qN D q

,29 where Csc is the capacitance of the space charge layer and A is the

electrode area, Vapp is the electrode potential, k is Boltzmann’s constant, ND is the


Page 14 of 27

Page 15 of 27 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


charge carrier density, T is the absolute temperature, ε is the static dielectric of Ta3N5 (~110)5, εo is the permittivity of free space, and q is the electronic charge. MS plots with positive slope imply that these Ta3N5 with or without Sc doping are n-type conducting semiconductor. The 5% Sc-Ta3N5 nanorods (Vfb = -0.2 VRHE) demonstrates cathodic shift of Vfb by about 100 and 300 mV relative to the Ta3N5 nanorods (Vfb = -0.1 VRHE ) and Ta3N5 particles (Vfb = 0.1 VRHE), respectively. Given that the flux-assisted nitridation route is beneficial to the reduction of VN defects, which can induce localized deep states.27,30,31 Therefore, the cathodic shift of Vfb of both Ta3N5 and Sc-doped Ta3N5 nanorods may originate from the decrease of surface VN defects, thus relaxing the undesired surface Fermi level pinning effects. We therefore measured the pseudo-Vfb under quasi-equilibrium conditions by carrying out the open circuit voltage measurements under dark (OCVdark) and illumination (OCVlight).12,32 The cathodic shift of OCVlight was obvious for both Ta3N5 and Sc-doped Ta3N5 nanorods, while the open circuit photovoltage (OCP) increased from the 250 mV for Ta3N5 particles, to 270 mV for Ta3N5 nanorods, and up to 320 mV for 5% Sc-Ta3N5 (Fig. 5d). The OCVlight presented the quasi-Fermi level, which is determined by the electronic structures of the Ta3N5, provided a reasonable comparison of how the true Fermi level changes. Photovoltage of photoelectrode means the change of internal driving force for charge separation from dark to irradiation. Therefore, the increase in OCP would mean that a relaxed surface Fermi level pinning strengthened built-in potential, thus contributing to the enhanced charge separation and the decreased charge recombination. The carrier densities in both 5%



Sc-Ta3N5 and Ta3N5 nanorods determined by MS equation were comparable because they present the almost completely same particle features except Sc doping. The carrier densities in both 5% Sc-Ta3N5 (5.87×1019 cm-3) and Ta3N5 (2.21×1019 cm-3) nanorods determined by MS equation were comparable because they present the almost completely same particle features except Sc doping. Theoretical calculations have revealed that ON defects in Ta3N5 add localized donor energy level below the conduction band,27,28,30,31 which contribute to a higher carrier density. The increased carrier density after Sc doping would originate from the substitution of O2- for N3compensating the charge difference between Sc3+ and Ta5+. The charge transport properties of the photoanodes were evaluated using electrochemical impedance spectroscopy (EIS) under light illumination (Figure S6). An equivalent circuit model analysis (Table S3) manifests that the impedance at the semiconductor-electrolyte interface (Rlight) for Ta3N5 nanorods is much smaller than that of Ta3N5 particles, and the Rlight was further reduced for 5% Sc-Ta3N5, confirming the elevated charge separation and transfer process by increase of ON electron donor from Sc doping. Correspondingly, the depletion layer capacitance (Clight) exhibits a trend of 5% Sc-Ta3N5 ＞ Ta3N5 nanorods ＞ Ta3N5 particles, in good agreement with their electrochemical surface area (Figure S7). This result further demonstrated that carrier density of Ta3N5, contributing by the ON electron donor, is indeed increased after Sc doping. In addition, the charge separation efficiency was evaluated based on photocurrent test using Na2SO3 as hole sacrificial reagent (Figure S8). The 5% Sc-Ta3N5 nanorods photoanode shows a significant increase in charge separation


Page 16 of 27

Page 17 of 27 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


efficiency if compared to Ta3N5 nanorods and particles, confirming the effectiveness from both reduction of VN and increment of ON electron donor contributing to improving bulk carrier separation and transfer. According to these experimental evidences, we can conclude the negative shift in the Vfb of Sc-doped Ta3N5 nanorods photoanode is from reduced trap states and increased conductivity. Co(OH)x modification of different photoanodes. Most importantly, after loading Co(OH)x as an oxygen-evolving reaction (OER) catalyst for reducing the OER overpotential and improving the stability of photocatalyst, the photocurrent onset potential further cathodically shifted to 0.4 VRHE for 5% Sc-Ta3N5/Co(OH)x, which is more than 27.3% and 38.5% increase compared to the Ta3N5 nanorods/Co(OH)x (0.55 VRHE) and Ta3N5 particles/Co(OH)x (0.65 VRHE), respectively (Figure 6a). Until now, the photocurrent onset potential at 0.4 VRHE was the lowest value, surpassing all previous reported Ta3N5 photoanodes (Table S1). Under both positive and negative-going potential sweep, the 5% Sc-Ta3N5/Co(OH)x exhibited the almost completely coincident current-potential paths (Figure S9). Morever, the photocurrent of 5% Sc-Ta3N5/Co(OH)x photoanode at 0.4, 0.45, 0.5 and 0.55 VRHE was further confirmed by steady-state photocurrent curve, respectively. As shown in Figure S10, a nearly constant photocurrent was achieved under i-t test during the light irradiation, agree well with the photocurrent value from the photocurrent-potential curve (Figure 6a). These evidences indicated that the photocurrent density was indeed not from oxidation of Co species, even if the applied potential is at the onset potential of 0.4 VRHE.10 The current density reached 4.9 mA cm-2 at 1.23 VRHE for the 5%



Sc-Ta3N5/Co(OH)x, which showed 53.1% and 32.4% improvement compared to Ta3N5 particles/Co(OH)x (2.3 mA cm-2) and Ta3N5 nanorods/Co(OH)x (3.7 mA cm-2), respectively. The faraday efficiency for 5% Sc-Ta3N5/Co(OH)x photoanode for H2 and O2 evolved at 1.23 VRHE was 85% and 77%, respectively, confirming that anodic photocurrent mainly originated from the water oxidation reaction (Figure 6b). The ratio of H2:O2 produced was 2.21:1. The slight difference from the stoichiometric 2:1 of H2:O2 for water splitting was probably attributed to the partial oxidization of Ta3N5 by photogenerated holes.7

Figure 6 (a) Current-potential curves tested under chopped AM 1.5G simulated sunlight. The inset shows the photocurrent onset potentials. (b) The oxygen and hydrogen evolved from 5% Sc-Ta3N5/Co(OH)x nanorods photoanode. The solid lines mean the theoretical amount of O2 and H2 evolution with 100% faraday efficiency. (c) IPCE dependence on the wavelength for the 5% Sc-Ta3N5/Co(OH)x nanorods


Page 18 of 27

Page 19 of 27 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


photoanode. (d) HC-STH for 5% Sc-Ta3N5/Co(OH)x and Ta3N5 particles/Co(OH)x photoanodes.

In addition to the reducing surface defects contributing to the high PEC activity of Ta3N5 nanorods, one-dimensional nanostructures have stimulated much interest in photoelectrochemical water splitting due to it results in enhanced carrier collection, and improved light absorption. In our case, the Ta3N5 nanorods electrode remains interconnected loose framework structure if compared to Ta3N5 particles electrode (Figure S11). Indeed, the relative electrochemical surface areas of Ta3N5 nanorods and 5% Sc-Ta3N5 photoanode are 2.3 and 3.0 times higher relative to that of Ta3N5 particles, respectively (Figure S7b), due to the increased interface region between semiconductor and electrolyte. More importantly, such unique structures allow carriers only need to migrate to the semiconductor/electrolyte junction across the radius of the nanorods (b, c axes) rather than transport along the longitudinal direction (a axis). Especially, theoretical studies have indicated that orthorhombic Ta3N5 is a highly anisotropic semiconductor with the hole and electron effective masses along a-axis direction are much larger than along b or c-axis direction.17,18 This means that to obtain Ta3N5 single crystal with a-axis preferential growth to keep sufficient light absorption but allow carriers easily migrate to the surface along b or c-axis direction would be a good choice to minimize the bulk recombination and negatively shift the photocurrent onset potential. The charge transport process was confirmed by comparing photocurrents achieved with front-side (Jfront, semiconductor side) and back-side illumination (Jback, FTO side) for Ta3N5 nanorods, Sc-doped Ta3N5



nanorods and Ta3N5 particles photoanode. Generally, Jfront/Jback < 1 and Jfront/Jback >1 indicate the electron transport limit and hole transport limit, respectively.25 Interestingly, at applied potential ranging from 0.2 to 1.0 VRHE, we observed that the Jfront was very close to Jback (Figure S12), whether for the Ta3N5 nanorods or the 5% Sc-Ta3N5 nanorods, although the Sc-doped Ta3N5 nanorods showed an increase in both Jfront and Jback. However, an indication of poor electron transport being a limiting factor for electron-hole separation in Ta3N5 particles was demonstrated by Jfront/Jback< 1 due to the serious surface recombination. It is suggesting that the as-fabricated Ta3N5 nanorods photoanodes have excellent inherent charge transport properties and suppress surface recombination. Figure 6c shows the incident photon-to-current conversion efficiency (IPCE) action spectra for the 5% Sc-Ta3N5/Co(OH)x nanorods photoanode. The IPCEs started to increase around 600 nm, conformed well to the light absorption spectra of Ta3N5 in Figure 4b. The IPCE values at 0.6 VRHE and 1.23 VRHE were ca. 2-10% and 40-55% in the range of 560-400 nm, respectively. Integrating IPCE data at 1.23 VRHE with the standard solar spectrum (Figure S13), gives a calculated photocurrent of ~5.1 mA cm-2, consistent with the value in Figure 6a. The HC-STH efficiency of 5% Sc-Ta3N5 nanorods photoanode was demonstrated in Figure 6d. The maximum HC-STH conversion efficiency was 0.82% at 0.9 VRHE, which surpasses the most previous observations (Table S1). The maximum applied bias photon-to-current efficiency (ABPE) measured from two-electrode system was calculated as 0.5% (Figure S14). After 2 h of illumination at 0.9 VRHE, the photocurrent of 5% Sc-Ta3N5/Co(OH)x


Page 20 of 27

Page 21 of 27 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


photoanode decayed by 40% (Figure S15). This performance decay could be attributed to the N anions on the surfaces of Ta3N5 were oxidized with photogenerated holes to generate surface oxides, which suppressed the charge transfer across the photoanode/electrolyte interface. Surface modification may be a potential way to increase the stability of nitride photoanodes.13

4. Conclusions In summary, the high-quality Sc doped Ta3N5 single-crystal nanorods can be synthesized by Na2CO3 flux assistance in flowing NH3 at 950 oC. After depositing Co(OH)x as co-catalyst, our photoanode exhibits a record water oxidation onset potential of 0.4 VRHE, resulting in 4.9 mA cm-2 at 1.23 VRHE with maximum HC-STH of 0.82%. Significantly, the reasons were understood as the improved crystalline quality, decreased negative effect of electronic structure anisotropy and cathodically shifted flat band potential owing to the morphological and electronic structure dual modulation. Our results contributed to understanding the nature of high water oxidation onset potential requirement for Ta3N5 photoelectrode and suggested that controlling crystal growth and element doping are effective strategy to developing high-activity nitrides-based semiconductor photoanodes.

ASSOCIATED CONTENT Supporting Information



Page 22 of 27

The Supporting Information is available free of charge on the ACS Publications website at DOI: XXXXX Photocatalyst

and

Photoelectrochemical

characterizations,

Theoretical

calculations. Comparison of our results with recent typical Ta3N5-based photoanodes (Table S1), Elemental content in various Ta3N5 photocatalysts measured by EDS and XPS (Table S2). Additional SEM image, HRTEM images, defect formation energies, current-potential curves, DFT calculations, EIS Nyquist plots, relative electrochemically active area, charge-separation efficiencies, steady-state photocurrent and applied bias photon-to-current efficiency.

AUTHOR INFORMATION Corresponding Author *E-mail: [email protected]

NOTES The authors declare no competing ﬁnancial interest. ACKNOWLEDGMENTS This work was supported primarily by the National Natural Science Foundation of China (51572121, 21603098, and 21633004), the Natural Science Foundation of Jiangsu Province (BK20151265, BK20151383, and BK20150580), the for

Outstanding

PhD

candidate

of

program

B

Nanjing University (201702B084), the

Postdoctoral Science Foundation of China (2017M611784), and the Fundamental Research Funds for the Central Universities (021314380133 and 021314380084). REFERENCES [1] Grätzel, Photoelectrochemical Cells. Nature, 2001, 414, 338-344.


Page 23 of 27 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


[2] Blankenship, R. E.; Tiede, D. M.; Barber, J.; Brudvig, G. W.; Fleming, G.; Ghirardi, M.; Gunner, M. R.; Junge, W.; Kramer, D. M.; Melis, A.; Moore, T. A.; Moser, C. C.; Nocera, D. G.; Nozik, A. J.; Ort, D. R.; Parson, W. W.; Prince, R. C.; Sayre, R. T. Comparing Photosynthetic and Photovoltaic Efficiencies and Recognizing the Potential for Improvement. Science, 2011, 332, 805-809. [3] Hong, W. T.; Risch, M.; Stoerzinger, K. A.; Grimaud, A.; Suntivich, J.; Yang, S. H. Toward the Rational Design of Non-Precious Transition Metal Oxides for Oxygen Electrocatalysis. Energy Environ. Sci., 2015, 8, 1404-1427. [4] Walter, M. G.; Warren, E. L.; McKone, J. R.; Boettcher, S. W.; Mi, Q.; Santori, E. A.; Lewis, N. S. Solar Water Splitting Cells. Chem. Rev., 2010, 110, 6446-6473. [5] Li, Y.; Zhang, L.; Torres-Pardo, A.; Gonzalez-Calbet, J. M.; Ma, Y.; Oleynikov, P.; Terasaki, O.; Asahina, S.; Shima, M.; Cha, D.; Zhao, L.; Takanabe, K.; Kubota, J.; Domen, K. Cobalt Phosphate-Modified Barium-Doped Tantalum Nitride Nanorod Photoanode with 1.5% Solar Energy Conversion Efficiency.

Nat. Commun., 2013, 4, 2566. [6] Feng, J.; Luo, W.; Fang, T.; Lv, H.; Wang, Z.; Gao, J.; Liu, W.; Yu, T.; Li, Z.; Zou, Z. Highly Photo-Responsive LaTiO2N Photoanodes by Improvement of Charge Carrier Transport Among Film Particles. Adv. Funct. Mater., 2014, 24, 3535-3542. [7] Li, M.; Luo, W.; Cao, D.; Zhao, X.; Li, Z.; Yu, T.; Zou, Z. A Co-Catalyst-Loaded Ta3N5 Photoanode with a High Solar Photocurrent for Water Splitting Upon Facile Removal of the Surface Layer. Angew. Chem. Int. Ed., 2013, 52, 11016-11020. [8] Chun, W.; Ishikawa, A.; Fujisawa, H.; Takata, T.; Kondo, J. N.; Hara, M.; Kawai, M.; Matsumoto, Y.; Domen, K. Conduction and Valence Band Positions of Ta2O5, TaON, and Ta3N5 by UPS and Electrochemical Methods. J. Phys. Chem.

B, 2003, 107, 1798-1803. [9] Liu, G.; Ye, S.; Yan, P.; Xiong, F.; Fu, P.; Wang, Z.; Chen, Z.; Shi, J.; Li, C. Enabling an Entegrated Tantalum Nitride Photoanode to Approach the ACS Paragon Plus Environment


Theoretical Photocurrent Limit for Solar Water Splitting. Energy Environ. Sci. 2016, 9, 1327-1334. [10] Seo, J.; Takata, T.; Nakabayashi, M.; Hisatomi, T.; Shibata, N.; Minegishi, T.; Domen, K. Mg-Zr Cosubstituted Ta3N5 Photoanode for Lower-Onset-Potential Solar-Driven Photoelectrochemical Water Splitting. J. Am. Chem. Soc., 2015,

137, 12780-12783. [11] Wang, L.; Zhou, X.; Nguyen, N. T.; Hwang, I.; Schmuki, P. Strongly Enhanced Water Splitting Performance of Ta3N5 Nanotube Photoanodes with Subnitrides.

Adv. Mater., 2016, 28, 2432-2438. [12] He, Y.; Thorne, J. E.; Wu, C. H.; Ma, P.; Du, C.; Dong. Q.; Wang, D. What Limits the Performance of Ta3N5 for Solar Water Splitting? Chem, 2016, 1, 640-655. [13] Zhong, M.; Hisatomi, T.; Sasaki, Y.; Suzuki, S.; Teshima, K.; Nakabayashi, M.; Domen, K. Highly Active GaN-Stabilized Ta3N5 Thin-Film Photoanode for Solar Water Oxidation. Angew. Chem. Int. Ed., 2017, 56, 1-6. [14] Pei, L.; Xu, Z.; Zhi, Z.; Zhu, H.; Yan, S.; Zou, Z. Mg-Doped Ta3N5 Nanorods Coated with a Conformal CoOOH Layer for Water Oxidation: Bulk and Surface Dual Modification of Photoanodes. J. Mater. Chem. A, 2017, 5, 20439-20447. [15] Zhen, C.; Chen, R.; Wang, L.; Liu, G.; Cheng, H. M. Tantalum (Oxy)Nitride Based Photoanodes for Solar-Driven Water Oxidation. J. Mater. Chem. A, 2016, 4, 2783-2800. [16] Fu, G.; Yan, S.; Yu, T.; Zou, Z. Oxygen Related Recombination Defects in Ta3N5 Water Splitting Photoanode. Appl. Phys. Lett., 2015, 107, 171902. [17] Morbec, J. M.; Narkeviciute, I.; Jaramillo, T. F.; Galli, G. Optoelectronic Properties of Ta3N5: A Joint Theoretical and Experimental Study. G. Phys. Rev.

B, 2014, 90, 155204. [18] Morbec, J. M.; Galli, Charge Transport Properties of Bulk Ta3N5 from First Principles. G. Phys. Rev. B, 2016, 93, 035201.


Page 24 of 27

Page 25 of 27 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


[19] Li, K.; Xue, D. Estimation of Electronegativity Values of Elements in Different Valence States. J. Phys. Chem. A, 2006, 110, 11332-11337. [20] Hwu, S. J.; Ziebarth, R. P.; Winbush, S. V.; Ford, J. E.; Corbett, J. D. Synthesis and Structure of Double-Metal-Layered Scandium, Yttrium, and Zirconium Chloride Carbides and Nitrides, M2Cl2C and M2Cl2N. Inorg. Chem., 1986, 25, 283. [21] Shulga, Y. U. M.; Troitskii, V. N.; Aivazov, M. I.; Borodko, Y. U. G. X-Ray Photo-Electron Spectra of the Mononitrides of Sc, Ti, V and Cr. Zh. Neorg.

Khim., 1976, 21, 2621-2624. [22] Ritala, M.; Kalsi, P.; Riihelä, D.; Kukli, K.; Leskelä, M.; Jokinen. J.; Sivula, K. Controlled Growth of TaN, Ta3N5, and TaOxNy Thin Films by Atomic Layer Deposition. Chem. Mater., 1999, 11, 1712-1718. [23] Henderson. S. J.; Hector, A. L. Structural and Compositional Variations in Ta3N5 Produced by High-Temperature Ammonolysis of Tantalum Oxide. J. Solid State

Chem., 2006, 179, 3518-3524. [24] Wang, J.; Feng, J.; Zhang, L.; Li, Z.; Zou, Z. Role of Oxygen Impurity On the Mechanical

Stability

and

Atomic

Cohesion

of

Ta3N5

Semiconductor

Photocatalyst. Phys. Chem. Chem. Phys., 2014, 16, 15375-15380. [25] Kim, T. W.; Ping, Y.; Galli, G. A.; Choi, K. S. Simultaneous Enhancements in Photon Absorption and Charge Transport of Bismuth Vanadate Photoanodes for Solar Water Splitting. Nat. Commun., 2015, 6, 8769. [26] Wang, J.; Ma, A.; Li, Z.; Jiang, J.; Feng. J.; Zou, Z. Effects of Ba-O Codoping on the Photocatalytic Activities of Ta3N5 Photocatalyst: a DFT Study. RSC Adv., 2014, 4, 55615. [27] Jing, T.; Dai, Y.; Ma, X. C.; Wei, W.; Huang, B. B. Effects of Intrinsic Defects and Extrinsic Doping on the Electronic and Photocatalytic Properties of Ta3N5.

RSC Adv., 2015, 5, 59390-59397. [28] Xie, Y.; Wang, Y.; Chen, Z.; Xu, X. Role of Oxygen Defects on the Photocatalytic Properties of Mg-Doped Mesoporous Ta3N5. ChemSusChem, 2016, 9, 1403-1412. ACS Paragon Plus Environment


Page 26 of 27

[29] Gelderman, K.; Lee, L.; Donne, S. W. Flat-Band Potential of a Semiconductor: Using the Mott-Schottky Equation. J. Chem. Educ., 2007, 84, 685. [30] Harb, M.; Sautet, P.; Nurlaela, E.; Raybaud, P.; Cavallo. L.; Domen, K. Tuning the

Properties

Photocatalysts

of by

Visible-Light-Responsive Non-Stoichiometric

Tantalum

Compositions:

a

(Oxy)

Nitride

First-Principles

Viewpoint. Phys. Chem. Chem. Phys., 2014, 16, 20548-20560. [31] Wang, Y.; Zhu. D.; Xu, X. Zr Doped Mesoporous Ta3N5 Microspheres for Efficient Photocatalytic Water Oxidation. ACS Appl. Mater. Interfaces, 2016, 8, 35407-35418. [32] Chen, Z.; Jaramillo, T. F.; Deutsch, T. G.; Kleiman-Shwarsctein, A.; Forman, A. J.; Gaillard, N.; McFarland, E. W. Accelerating Materials Development for Photoelectrochemical Hydrogen Production: Standards for Methods, Definitions, and Reporting Protocols. J. Mater. Res., 2010, 25, 3-16.


Page 27 of 27 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Morphological and electronic structure dual modulation of Sc doped Ta3N5 nanorod is developed by fluxassisted oriented crystal growth route to significantly overcome the electric structure anisotropy, suppress surface recombination and tune the energy level alignment. As a result, the onset potential of Ta3N5 photoanode is unprecedentedly negatively shifted to 0.4 VRHE, and thereby affording a 0.82% half-cell solar-to-hydrogen efficiency. 35x15mm (300 x 300 DPI)


Oriented Growth of Sc-Doped Ta3N5 Nanorod Photoanode Achieving

Recommend Documents