Trade-off between Zr Passivation and Sn Doping on Hematite

Subscriber access provided by United Arab Emirates University | Libraries Deanship

Article

Trade-off between Zr passivation and Sn doping on hematite nanorod photoanodes for efficient solar water oxidation: Effects of a ZrO underlayer and FTO 2

Arunprabaharan Subramanian, Alagappan Annamalai, Hyun-Hwi Lee, Sun Hee Choi, Jungho Ryu, Jung Hee Park, and Jum Suk Jang ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.6b04528 • Publication Date (Web): 15 Jul 2016 Downloaded from http://pubs.acs.org on July 16, 2016

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 free 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 accessible to all readers and 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.

ACS Applied Materials & Interfaces 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 11

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 Materials & Interfaces

Trade-off between Zr Passivation and Sn Doping on Hematite Nanorod Photoanodes for Efficient Solar Water Oxidation: Effects of a ZrO2 Underlayer and FTO Deformation Arunprabaharan Subramaniana, Alagappan Annamalaia*, Hyun Hwi Leeb, Sun Hee Choib, Jungho Ryuc, Jung Hee Parka and Jum Suk Janga* a

Division of Biotechnology, Safety, Environment and Life Science Institute, College of Environmental and Bioresource Sciences, Chonbuk National University, Iksan 570-752, Republic of Korea. b Pohang Accelerator Laboratory (PAL), Pohang University of Science and Technology (POSTECH), Pohang 790-784, Republic of Korea. c Mineral Resources Research Division, Korea Institute of Geoscience and Mineral Resources (KIGAM), Daejeon 305-350, Korea.

ABSTRACT: Herein we report the influence of a ZrO2

underlayer on the PEC (photoelectrochemical) behavior of hematite nanorod photoanodes for efficient solar water splitting. Particular attention was given to the cathodic shift in onset potential and photocurrent enhancement. Akaganite (β-FeOOH) nanorods were grown on ZrO2 coated FTO (Fluorine doped tin oxide) substrates. Sintering at 800°C transformed akaganite to the hematite (α-Fe2O3) phase and induced Sn diffusion into the crystal structure of hematite nanorods from the FTO substrates and surface migration, shallow doping of Zr atoms from the ZrO2 underlayer. The ZrO2 underlayer treated photoanode showed better water oxidation performance compared to the pristine (α-Fe2O3) photoanode. A cathodic shift in the onset potential and photocurrent enhancement was achieved by surface passivation and shallow doping from the ZrO2 underlayer, along with Sn doping from the FTO substrate to the crystal lattice of hematite nanorods. The Zr-based hematite nanorod photoanode achieved 1 mA/cm2 at 1.23 VRHE with a low turn-on voltage of 0.80 VRHE. Sn doping and Zr passivation, as well as shallow doping, were confirmed by XPS, Iph and M-S plot analyses. Electrochemical impedance spectroscopy revealed that the presence of a ZrO2 underlayer decreased the deformation of FTO, improved electron transfer at the hematite/FTO interface and increased charge-transfer resistance at the electrolyte/hematite interface. This is the first systematic investigation of the effects of Zr passivation, shallow doping and Sn diffusion on hematite nanorod photoanodes through application of a ZrO2 underlayer on the FTO substrate.

KEYWORDS: Akaganite, cathodic shift, passivation, Shallow doping, Onset potential, Photoelectrochemical cells

1. INTRODUCTION The conversion of solar energy to chemical energy in the form of hydrogen and oxygen is needed due to depletion of fossil fuels. Photoelectrochemical (PEC) water splitting is an effective method of converting solar energy to chemical energy.1 Several materials have been examined for PEC water splitting; these include TiO2,2 WO3,3 BiVO4,4 ZnO5 and Fe2O3.6 Among them, hematite is a promising candidate because it has a narrow band gap, absorbance in the visible region, chemical stability and abundance in the earth’s crust, but also has several disadvantages, such as a short hole diffusion length, poor conductivity and charge recombination7. Several methods of overcoming these limitations have been evaluated. 1-D nanorods effectively reduce short hole diffusion.8 The poor conductivity of the hematite photoanode was rectified by introducing foreign atoms into the hematite lattice (doping).9 Surface treatment of hematite photoanodes results in increased conductivity and a significant enhancement of the photocurrent.10 Another problem is charge recombination, which decreases the photo activity of hematite photoanodes. Charge recombination mostly takes place at the interface between TCO (transparent conducting oxide)/photoanode material and TCO/electrolyte.11 This problem occurs because a dead layer exists at the interface between the TCO and photoanode, which can limit photoanode activity. A simple method of avoiding this problem is application of a metal oxide underlayer before growth of the photoanode material. This oxide underlayer decreases the negative effect on device performance, which reduces charge recombination and pro-

ACS Paragon Plus Environment


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

motes water oxidation.12 Recently, researchers have focused on water splitting using photogenerated charge carriers with minimum external bias. For n-type semiconductors, illumination with energy equal to or higher than the band gap of the material results in minority hole carriers driving OER and majority electron carriers driving HER at the electrode/electrolyte interface.13 The potential at which the above reaction starts is termed the photocurrent onset potential. For low applied bias photoelectrochemical water splitting, a material that has a cathodic shift in onset potential (low turn on voltage) and higher increment in photocurrent at the water oxidation potential is needed. The surface of hematite contains unwanted surface states caused by the Fe3+/Fe2+ redox couple in oxygen-deficient regions that trap holes or electrons on the surface, which causes recombination and lower oxygen reaction kinetics during the water oxidation reaction.14 Surface recombination at trapping sites causes an over-potential for the water oxidation reaction to take place. These problems can be overcome by applying a thin oxide layer on the hematite surface, which suppresses the surface states and enhances PEC performance by passivation. 15,16,17,18 Alternatively, the presence of an underlayer suppresses the charge recombination at the interfaces and enhances the PEC performance of hematite photoanodes. Hisatomi et al. investigated the performance of a Ga2O3 underlayer on ultrathin hematite in terms of water-splitting performance with a cathodic shift in onset potential by 0.2 V.19 A Nb2O5 underlayer treated ultrathin hematite photoanode showed enhanced photocurrent due to passivation of the FTO substrate.20 Previously, we developed TiO212 and GO21 underlayers for hematite photoanodes to enhance the photocurrent density. The presence of a TiO2 underlayer decreases the surface recombination and provides Ti4+ doping, while a GO underlayer decreases the micro strain and minimizes defects at the interface, leading to an increment in photocurrent. In this work, we modified a hematite nanorod photoanode by introducing a ZrO2 underlayer. This is the first report of a ZrO2 underlayer for hematite nanorods to provide a cathodic shift in onset potential as well as enhancement of photocurrent. Akaganite nanorods were grown on ZrO2-modified FTO substrates. The asprepared photoanodes were subjected to hightemperature sintering at 800°C, which caused a phase transformation from akaganite to hematite, penetration of Sn atoms from FTO substrates and Zr migration, and shallow doping of Zr atoms from the ZrO2 underlayer. As the concentration of the ZrO2 underlayer increased, Sn diffusion from the FTO substrate decreased. An equilibrium between doping of Sn atoms from the FTO substrate into the hematite lattice and Zr atoms from the ZrO2 underlayer was maintained. Migration of Zr atoms

Page 2 of 11

from the ZrO2 underlayer to the top surface of hematite nanorods was confirmed by XPS data. This causes passivation, which eventually decreases the surface defects that led to a cathodic shift in onset potential by 0.07 V. 2. EXPERIMENTAL SECTION Fluorine-doped tin oxide (FTO) substrates were cleaned with distilled water, ethanol and acetone. For the ZrO2 underlayer, the cleaned FTO plates (2.5 cm × 1 cm) were coated with 0.7, 1.75, 3.5 and 5.25 mM Zirconyl nitrate solution by spin-coating (2500 rpm for 25 s), and were then heated at 200°C for 30 min. Hematite nanorods on FTO plates and FTO plates treated with zirconium precursor were prepared by a simple hydrothermal method.6 The FTO and Zr-treated FTO plates were placed in a vial containing 10 mL of a solution of 0.4 g FeCl3.6H2O and 0.85 g NaNO3 at pH 1.5 (pH adjusted with HCl). The hydrothermal reaction was conducted at 100°C for 6 h. After cooling to room temperature, the FTO plates were rinsed several times with distilled water. Annealing at 800°C for 10 min was carried out for the phase transition from β– FeOOH to generate pure α–Fe2O3 and activation of hematite nanorods by Sn doping into the crystal structure (Scheme 1). The as-prepared photoanodes were denoted as pristine (α-Fe2O3), 0.7 mM Zr, 1.75 mM Zr, 3.5 mM Zr, and 5.25 mM Zr-α-Fe2O3. The crystallinity and preferential growth orientation of the ZrO2 underlayer coated α-Fe2O3 photoanodes were investigated by two-dimensional x-ray diffraction (2D-XRD). XRD measurements were performed at the 5A Materials Science XRS beamline of the Pohang Light Source II (PLS-II) in Korea at an x-ray wavelength of 1.008 Å. Diffuse reflectance ultraviolet-visible absorption spectra of the films were collected using a Shimadzu UV-Vis spectrophotometer (UV-2600). Field emission scanning electron microscopy (FESEM, Zeiss) was used to analyze the morphology of the photoanodes. The presence of Zr at the interface between FTO and hematite was confirmed by transmission electron microscopy (JEM2100F HR, JEOL) using a focused ion beam (Helios NanoLab, FEI). The oxidation state and binding energy of elements in the as-prepared photoanodes were confirmed using an X-ray photoelectron spectrometer (Thermo Scientific). Extended X-ray absorption fine structure (EXAFS) experiments were carried out at the 7D beamline of the Pohang Accelerator Laboratory (PLS-II, 3.0 GeV, 400 mA). The synchrotron radiation was monochromatized using a Si (111) double crystal monochromator. At room temperature, the spectra for the Fe K-edge (E0=7112 eV) were obtained in fluorescence mode. The incident beam was detuned by 20% for the Fe K-edge to minimize contamination of higher harmonics, and the intensity was monitored using a N2filled IC SPEC ionization chamber. The fluorescence signal from the sample was measured using a passivated

2


Page 3 of 11

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


implanted planar silicon (PIPS) detector. During the measurements, helium was continuously flowed into the sample chamber so that air scattering of the fluorescence signal was minimized to give a higher S/N ratio. AHENA and ARTEMIS in the IFEFFIT software suite were used to analyze the obtained data for the local structure of Fe in ZrO2 underlayer-treated α-Fe2O3 photoanodes.22 A standard for fitting experimentally derived radial structural functions was generated with FEFF 9 code23 using the known α-Fe2O3 structure.24 Thickness of the ZrO2 underlayer measured by ellipsometry (J. A. Woollam Co. RC2 Ellipsometer). Photoelectrochemical characterizations were performed using an Ivium CompactStat potentiostat in 1 M NaOH electrolyte solution (pH 13.6). A three-electrode cell configuration was used in this Figure 1. (a) X-ray diffraction patterns of pristine and Zrα-Fe2O3 photoanodes. (b) XRD patterns near the hematite (104) and (110) peaks. (c) The variations of intensity ratio (black line) of the (104) peak and (110) peak as a function of the zirconium precursor concentration.

Scheme 1. Experimental procedure for the preparation of pristine and α-Fe2O3 photoanodes grown on FTO treated with a zirconium precursor. experiment, with the fabricated material acting as the working electrode, Ag/AgCl (saturated KCl) as the reference electrode, and platinum coil as the counter electrode. The potential was reported against the reversible hydrogen electrode (RHE) by equation (1). ERHE = EAgCl + 0.059 pH + EoAgCl (EoAgCl = 0.1976 V at 25 °C) (1) 3. RESULTS AND DISCUSSION Figure 1 (a) shows the X-ray diffraction patterns of pristine and α-Fe2O3 nanorods grown on FTO treated with a Zr precursor. The crystal structure of pristine and Zr-αFe2O3 nanorods shows an α-Fe2O3 phase. The XRD patterns of the as-prepared samples were well matched with SnO2 (SnO2: JCPDS # 77-0452) on FTO substrates along with hematite (α-Fe2O3: JCPDS #33-0664) diffraction peaks, and no impurity peak related to Zr was observed. Diffraction peaks of (110) and (300) were observed for all of the photoanodes. A highly conductive plane (110) was displayed by both pristine and Zr-αFe2O3 nanorod photoanodes.25 The main crystalline phase was preserved after ZrO2 underlayer treatment. However, introduction of the ZrO2 underlayer caused slight and distinct changes in the relative intensities of the (110) and (104) peaks (Figure 1 (b) and (c)).

To investigate the effect of ZrO2 underlayer treatment on the preferential growth orientation of hematite nanorod photoanodes, 2D-XRD analysis was conducted (Figure 2 (a)-(d)). All of the 2D-XRD patterns were collected in quasi-symmetric reflection mode at Qz~2.4 Å-1. Figure 2 (e) and 2 (f) show the circular line cuts of the hematite (104) and (110) peaks, respectively. In Figure 2 (f), the maximum intensity of the α-Fe2O3 sample at 0° represents the preferential growth of the (110) plane. Treatment with the ZrO2 underlayer had little effect on the intensity distribution of the (110) and(104) planes. Therefore, neither the crystallinity nor the growth orientation was corrupted by the ZrO2 underlayer treatment. The optical properties of the as-prepared photoanodes were investigated using UV-Vis spectroscopy (Figure S1). Band gap values were calculated by extrapolating the curve obtained from the Tauc plot. Pristine and Zr-αFe2O3 nanorod photoanodes displayed similar band gaps. The presence of a ZrO2 underlayer did not change the optical properties of the pristine photoanode, which indicates shallow doping of Zr on hematite nanorods after ZrO2 underlayer treatment. Thus, all of the photoanodes displayed band gaps of ~2.1 eV. Figure 3 shows the morphology of pristine and Zr-α-Fe2O3 nanorod photoanodes. Nanorods were uniformly grown on the FTO substrates, as reported previously.26 There was no significant change in the length of nanorods after introduction of the ZrO2 underlayer. The pristine photoanode was ~300 nm in length. FESEM images of the Zr-α-Fe2O3 nanorod photoanodes were almost identical to those of the pristine photoanode,27 and

3



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

Page 4 of 11

Figure 4. FIB-TEM cross-sectional image of 1.75 mM Zr-α-Fe2O3 photoanode. The corresponding EDS spectrum shows the weight percentages of the elements. Figure 2. (a)–(d) 2D XRD images of pristine and Zr-αFe2O3 photoanodes. (e)-(f) The circular line cuts passing through the hematite (104) and (110) peaks, in which zero degrees corresponds to the surface normal direction.

their morphologies were preserved despite the growth of hematite nanorods on FTO treated with the Zr precursor.27 A higher concentration of zirconium underlayer hindered the growth of akaganite nanorods on the ZrO2modified FTO substrate. An FESEM image of α-Fe2O3 nanorods grown on FTO treated with 7 mM Zr (7 mM Zr-Fe2O3) is shown in Figure S2. Nanorods on some portions of the FTO surface were not homogeneously grown because of the inhomogeneous ZrO2 coating. Point EDS spectra were obtained for different regions of the FIB-treated photoanodes to detect Zr on the hematite nanorods.28 Zr was confirmed to be present at the interface between hematite and FTO, and the weight percent of Fe, Sn, O, and Zr are shown in Figure 4. After hightemperature annealing, a Zr content of ~0.2 weight percent was present at the interface between hematite and the FTO substrate. Zr atoms at the surface of hematite nanorods were not detected by FIB-TEM analysis because of the low concentration of the ZrO2 underlayer. XPS analysis showed the presence of Fe, Sn and O on the top surface of hematite nanorods. Indeed, Zr was also present (Figure 5), and the atomic percent of each element is shown in Table S2. Figure 5 shows XPS spectra of (a) Fe2p, (b) O1s, (c) Sn3d and (d) Zr3d. For all of the photoanodes, Fe2p was

Figure 3. FESEM images of pristine and Zr-α-Fe2O3 photoanodes: (a) pristine, (b) 0.7 mM Zr, (c) 1.75 mM Zr, (d) 3.5 mM Zr and (e) 5.25 mM Zr.

observed as Fe2p3/2 and Fe2p1/2 peaks at 711.2 eV and 724.2 eV, in agreement with previous reports. 29,30 The O1s peak was observed at 530.1 eV. Sn3d5/2 and Sn3d3/2 peaks were observed at 486.4 eV and 494.9 eV, respectively.10 For pristine photoanode, high-temperature sintering (800°C) induced Sn doping31 from the FTO substrate into the hematite nanorod crystal structure. For ZrO2 underlayer-modified photoanodes, Sn doping into the crystal structure, Zr passivation and shallow doping from the Zr underlayer to hematite nanorods occurred simultaneously during sintering at 800°C. For ZrO2 underlayer-treated photoanodes, Zr3d5/2 and Zr3d3/2 peaks were observed at 182.41 and 184.86 eV.32 However, Zr peaks were not observed with low concentrations of Zrα-Fe2O3. This result suggests that the migration of Zr atoms from the ZrO2 underlayer depends on the concentration of the Zr precursor and that only a portion of the Zr atoms migrate toward the top surface of α-Fe2O3 nanorods from the ZrO2 underlayer. The XPS results of 1.75 mM Zr precursor solution coated on FTO substrates sintered at 200°C for 30 min indicate formation of ZrO2. Two Zr3d peaks were observed as Zr3d3/2 and Zr3d5/2 at 184.26 eV and 182.41 eV (Figure S5), as reported previously33. An equilibrium was maintained between the amounts of dopants in the hematite nanostructure (Table S2). Sn diffusion into the hematite nanorod crystal structure from the FTO substrate was higher for pristine photoanodes, but the diffusion of Sn atoms decreased with increasing Zr precursor concentration, since the surface migration of Zr atoms interrupts Sn diffusion from the FTO substrate during the activation process. For instance, 5.25 mM Zr-α-Fe2O3 photoanode resulted in lower Sn diffusion compared to the pristine photoanode (Table S2). Figure 6 (a) illustrates the photocurrent density-voltage characteristics of the pristine and Zr-αFe2O3 photoanodes. The pristine photoanode showed a photocurrent density of 0.81 mA/cm2 at 1.23 VRHE with an onset potential of 0.87 VRHE. The 1.75 mM Zr-αFe2O3 photoanode had a photocurrent density of 1.0 mA/cm2 at1.23 VRHE with an onset potential of 0.8 VRHE. For photoanodes grown on FTO treated with Zr precursor, the onset potential was shifted to the cathodic side by 0.07 VRHE, and the photocurrent was enhanced 20% at 1.23 VRHE. Higher concentrations of zirconium de-

4


Page 5 of 11

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


creased the photocurrent of the photoanode. At >3.5 mM Zr, the photocurrent of the photoanode started to decrease, but the onset potential was unaffected. The 5.25 mM Zr-α-Fe2O3 photoanode had a lower photo current compared to the pristine photoanode and

a lower photocurrent compared to the pris tine photoanode, the current decay decreased by ~33

Figure 5. XPS spectra of the pristine and Zr-α-Fe2O3 photoanodes. (a) Fe2p, (b) O1s, (c) Sn3d and (d) Zr3d. exhibited a cathodic shift. This result is useful for solar water splitting applied with a low bias. The migration (passivation) as well as shallow doping of Zr atoms on hematite nanorods from the ZrO2 underlayer occurs along with diffusion of Sn from FTO substrate during high-temperature sintering. Passivation shows that the cathodic shift in onset potential and shallow doping results increment in photocurrent. At higher Zr precursor concentrations, passivation increases because of the relatively large amount of Zr atoms on the surface of the hematite nanorods, and conductivity decreases because of the relatively small amount of Sn atoms that diffuse into the hematite lattice. Therefore, the onset potential is unchanged, while the photocurrent decreases. The 1.75 mM Zr-α-Fe2O3 photoanode encompasses the trade -off point between passivation and conductivity (Scheme 2). Figure 6 (b) shows the transient photocurrent of the pristine and Zr-α-Fe2O3 photoanodes at 1.23 VRHE. A positive current transient was observed for all of the photoanodes, due to accumulation of holes at the interface between electrolyte/electrode upon exposure to light.34 This leads to charge recombination at the electrolyte/electrode interface. In the case of the pristine photoanode, the photocurrent decay was 104 µA, while in the case of 1.75 mM and 5.25 mM Zr-α-Fe2O3 photoanodes, the photocurrent decay decreased. Charge recombination is decreased when a Zr passivation layer is present on the hematite nanorod surface. The 1.75 mM Zr-α-Fe2O3 photoanode had a current decay of ~59 µA. Although the 5.25 mM Zr-α-Fe2O3 photoanode exhibited

Figure 6. (a) Photocurrent density–voltage curves of pristine (α-Fe2O3) and Zr-α-Fe2O3 photoanodes under illumination conditions using 1 M NaOH (inset shows the onset potential shift of Zr-α-Fe2O3). (b) Photocurrent transient curves of pristine and Zr-α-Fe2O3 photoanodes measured at 1.23 VRHE. µA due to lower charge recombination at the electrolyte /electrode interface. Figure 7 (a) shows photocurrent density-voltage curves of the 1.75 mM Zr-α-Fe2O3 photoanode in 1 M NaOH and 1 M NaOH + 0.5 M H2O2 as electrolytes. The photocurrent and dark current in the graph are represented by solid and dashed lines, respectively. The onset potential was shifted to more negative in the presence of H2O2. The current transient became absolutely faradaic after addition of H2O2 to the NaOH electrolyte solution. H2O2 reduces surface recombination, thus movement of accumulated holes toward the electrolyte from the photoanode leads to water oxidation.35, 36 Figure 7 (b) shows the photocurrent transient measurement of a 1.75 mM Zr-α-Fe2O3 photoanode in various electrolytes under 1.0 VRHE and 1.23 VRHE. Photocurrent transient measurement in 1 M NaOH + 0.5 M H2O2 electrolyte solution exhibited an exact steady state faradaic current even at low applied potential. For the NaOH electrolyte solution, an anodic spike in the photocurrent

5



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

transient curve decreased at higher voltage and increased at lower voltage. When exposed to light, accumulated holes at the interface between electrolyte and electrode at lower voltage recombine without injection into the electrolyte, thus the anodic spike increases in 1 M NaOH electrolyte solution when a low potential is applied. The addition of H2O2 to the electrolyte reduces the recombination of photo generated electrons and holes, giving a steady state faradaic current for all photoanodes even at a low applied voltage.36 ZrO2 underlayer-treated photoanodes exhibited an almost steady-state faradaic photocurrent in 1 M NaOH electrolyte. When the Zr concentration was increased , the anodic spike in the photocurrent transient curve decreased because of lower recombination

Page 6 of 11

structure (EXAFS) probes local neighboring atoms around a specific metal element. Figure 8 displays Fourier-transformed spectra of EXAFS functions for Fe Kedges of the ZrO2 underlayer-treated substrate. Two peaks appear within a radial distance of 4 Å: the first peak at 0.8–2.2 Å and a second at 2.2–3.9 Å. The former peak is due to the two nearest Fe-O bonds, and the latter peak is contributed by Fe-M (M=Fe or Sn or Zr) and FeO bonds at a longer distance. As annealing at 800°C causes Zr passivation as well as shallow doping on hematite nanorods from the underlayer along with Sn diffusion into the α-Fe2O3 lattice from the FTO substrate, the peak at 2.2-3.9 Å is composed of extensively complicated scatterings and cannot be separated. The intensity of the peak at 0.8–2.2 Å decreased with increasing Zr precursor concentrations up to 3.5 mM, but returned to the level for α-Fe2O3 at 5.25 mM. This peak is invariant in terms of radial distance position with increasing Zr concentration, but it shifted to a lower distance with 5.25 mM Zr precursor. These observations are supported by the EXAFS fitting results in Table 1. The Debye-Waller factors for the pristine and Zr-α-Fe2O3 photoanodes were much smaller than that of α-Fe2O3 powder, indicating that the structural disorder is diminished for nanorods grown on the FTO substrate. As the concentration of Zr precursor was increased to 3.5 mM, this factor increased slightly. At 5.25 mM, the Debye-Waller factor decreases; furthermore, the two Fe-O bond distances also decreased by 0.2–0.3 Å compared to α-Fe2O3. It has been reported that Sn is a good doping element for hematite, because Sn4+ has a similar ionic radius and Pauling electronegativity to Fe3+.37 However, Zr4+ exhibits a larger ionic radius and much smaller electronegativity compared to Fe3+.38 Recalling that the ZrO2 underlayer reduces Sn diffusion from the FTO substrate, as shown in the XPS data, Zr passivation as well as shallow doping becomes more pronounced relative to Sn diffusion as the Zr concentration is increased. Below 5.25 mM, such an effect causes a slight change in local structure by reducing the Fe-O bond order. However, sufficient Zr passivation reduces Fe-O bond distances and enhances the Fe-O bond order, because a respectable amount of Zirconium passivation as well as shallow doping with small electronegativity could introduce more electrons into neighboring Fe3+ sites and recover the structural order to a level comparable to α-Fe2O3.

Figure 7. (a) Photocurrent density–voltage curves of a 1.75 mM Zr-α-Fe2O3 photoanode under dark and illumination conditions using 1 M NaOH and 1 M NaOH+0.5 M H2O2. (b) Photocurrent transient curves of a 1.75 mM Zr-α-Fe2O3 photoanode measured at 1.0 V and 1.23 VRHE using 1 M NaOH and 1 M NaOH+ 0.5 M H2O2 electrolytes.

of holes and electrons. Therefore, the onsetpotential of the Zr-α-Fe2O3 photoanode was shifted to the negative side by Zr passivation. Extended X-ray absorption fine

6


Page 7 of 11

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


concentration of zirconium precursor decreased the sheet resistance of the FTO substrate. The pho toanode prepared with 5.25 mM ZrO2 underlayer gave the lowest sheet resistance of 32.25 Ω. At higher concentrations of zirconium, the sheet resistance of the FTO substrate

Figure 8. k3-weighted Fourier transforms of EXAFS functions for Fe K-edges of pristine and Zr-α-Fe2O3 photoanodes.

Table 1. Structural parameters calculated from Fe Kedge EXAFS fits for the pristine and Zr-α-Fe2O3 photoanodes Sample Pristine

R1 (Å)a 1.96

Sample

RS (Ω)

Pristine

70.1

0.7mM Zr

51.2

1.75 mM Zr

53.8

3.25 mM Zr

36.5

32.2 5.25 mM Zr 1.97 0.7 mM Zr 1.96 1.75 mM Zr 1.97 3.25 mM Zr 1.94 5.25 mM Zr 1.96 Ref* a,b

R2 (Å)b 2.13 RCT1 (Ω) 12.8 2 10.2 4 15.5 4 15.3 8 7.75 2.14 2.13 2.14 2.10 2.13

σ2 (Å2)c 0.0027(8)

R-factord 0.0162

CPE-1 (F)

RCT2 (Ω)

CPE-2 (F)

4.70x10-5

226.7

4.58x10-6

3.47x10-5

250.4

1.70x10-6

2.17x10-5

265.2

1.06x10-6

2.75x10-5

341.3

9.09x10-7

1.09x10-5 419 0.0024(8) 0.0030(8) 0.0032(11) 0.0021(9) 0.0086(16)

1.10x10-5 0.0165 0.0172 0.0168 0.0269 0.0074

Fe-O bond distance (uncertainty

Trade-off between Zr Passivation and Sn Doping on Hematite

Recommend Documents