Toward Robust Photoelectrochemical Operation of Cuprous Oxide

Subscriber access provided by READING UNIV

Energy, Environmental, and Catalysis Applications

Toward Robust Photoelectrochemical Operation of Cuprous Oxide Nanowire Photocathodes Using Strategically-Designed Solution-Processed Titanium Oxide Passivation Coating Joo Sung Kim, Sung Woon Cho, Nishad G Deshpande, Young Been Kim, Young Dae Yun, Sung Hyeon Jung, Dong Su Kim, and Hyung Koun Cho ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.9b02727 • Publication Date (Web): 02 Apr 2019 Downloaded from http://pubs.acs.org on April 3, 2019

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 13 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

Toward Robust Photoelectrochemical Operation of Cuprous Oxide Nanowire Photocathodes Using Strategically-Designed SolutionProcessed Titanium Oxide Passivation Coating Joo Sung Kim,‡ Sung Woon Cho,‡ Nishad G. Deshpande,‡ Young Been Kim, Young Dae Yun, Sung Hyeon Jung, Dong Su Kim, and Hyung Koun Cho* School of Advanced Materials Science and Engineering, Sungkyunkwan University, 2066, Seobu-ro, Jangan-gu, Suwon-si, Gyeonggi-do, 16419, Republic of Korea * Corresponding author's email: [email protected]

ABSTRACT: To date, TiO2 films prepared by atomic layer deposition (ALD) are widely used to prepare Cu2O nanowires (NWs)based photocathodes with photoelectrochemical (PEC) durability as this approach enables conformal coating and furnishes chemical robustness. However, this common approach requires complicated interlayers and makes the fabrication of photocathodes with reproducible performance and long-term stability difficult. Although sol-gel based approaches have been well-established for coating surfaces with oxide thin-films, these techniques have rarely been studied for oxide passivation in PEC applications because the sol-gel coating methods are strongly influenced by surface chemical bonding and have been mainly demonstrated on flat substrates. As a unique strategy based on solution processing, herein, we suggest a creative solution for two problems encountered in the conformal coating of surfaces with oxide layers: i) how to effectively prevent corrosion of materials with hydrophilic surfaces by simply using a single TiO2 surface protection layer, instead of a complex multilayer structure, and ii) guaranteeing perfect chemical durability. Cu(OH)2 NW scan be easily prepared as an intermediate phase by anodization of Cu metal, where the former inherently possess a hydrophilic hydroxylated surface, and thus enables thorough coating with TiO2 precursor solutions. Chemically robust nanowires are then generated as the final product via phase transformation of Cu(OH)2 to Cu2O via sintering at 600 °C. The coated NWs exhibit excellent PEC properties and stable performance. Consequently, the perfect chemical isolation of the Cu2O NWs from the electrolyte allows remarkable PEC operation with the maintenance of the initial photocurrent for more than one day. Keywords:Cu2O; Cu(OH)2; TiO2; Nanowires; Photocathode; Photoelectrochemical (PEC) and Anodization. Introduction The depletion of fossil fuel reserves and environmental pollution associated with the use of this resource are driving factors behind the growing need for reliable and sustainable power sources. Solar energy and water reservoirs are prospectively the most efficient alternatives because they constitute the most abundant and the largest renewable energy sources on earth. One practical way to store a large amount of energy is to use chemical energy carriers such as fuels. The photoelectrochemical cell (PEC), which produces electrical energy or hydrogen from water using solar energy, is a truly promising system for meeting the emerging energy storage requirements. The p-type semiconductor cuprous oxide (Cu2O) is one of the most promising photocathode materials for the operation of PEC cells because it has an appropriate direct bandgap of 2.1 eV and can effectively absorb over a substantial range of the solar spectrum, and has a proper energy band position for hydrogen evolution.1 Several groups have studied Cu2O absorbers for use in low-cost thin-film solar cells, but the energy conversion efficiency of these absorbers is still unsatisfactory.2-3 On the other hand, the Grätzel group, one of the most distinguished research groups focusing on electrolyte-based photovoltaic application, selected Cu2O films as photocathode materials and reported favorable results.4-6 However, Cu2O still suffers from severe limitations

as an absorption layer for PEC operation. For example, the very short diffusion length of 150 nm induces recombination loss; thus, sufficient hydrogen cannot be generated from the electrolyte.7 To overcome such issues, the use of Cu2O NWs has been suggested. To maximize the PEC reaction between the electrolyte and Cu2O, this nanostructure can provide a massive increase in the contact area between the photoelectrode and electrolyte, and the distance for electron movement into the electrolyte is short. Additionally, this structure allows relatively high absorbance and stray radiation can be reabsorbed by other NWs.8 Owing to these advantages, layers of vertically aligned Cu2O NWs can be considered as ideal photo-absorption layers. However, Cu2O phase is chemically unstable due to active photo-corrosion effects.9-11 Because the redox potentials of Cu2O corresponding to reductive and oxidative corrosion are positioned within the bandgap, the photo-generated electrons or holes spontaneously participate in the corrosion process under visible-light irradiation. Thus, an additional protective surface layer that perfectly covers the Cu2O surface is required.12 Among the various surface protective films with adequate band alignment and chemical durability for photocathodes, TiO2 films are presently accepted as the best oxide material, where TiO2 plays the multiple roles of ideal electron transport and hole blocking, as well as chemical protection. Seger et al.13 reported that a sputtered TiO2 layer provided

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces 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

60h protection of a Si photocathode without significant degradation during photocatalytic H2 evolution. Moreover, for Si, GaAs, and GaP photoanodes, it was reported that thin atomic layer deposited (ALD) TiO2 coatings with inadequate band alignment surprisingly prevent photochemical corrosion and promote hole conduction via low tunneling resistance. More recently, amorphous TiO2 has even been utilized in protecting the photoelectrode layer.14,15Although there are different ways in order to achieve the TiO2 coatings, the use of TiO2 thin films coated onto surfaces via ALD has been preferentially recommended for achieving the desired photo-conversion efficiency and corrosion resistance. This approach utilizes a direct coating of single-layer TiO2 without intermediate layers, as the single layer provides sufficient protection despite being a few hundred nm thick. It is also reported and suggested that a thin TiO2 layer using ALD can provide slightly improved photovoltages and photocurrents instead of thick TiO2 coatings.16 Based on the above studies, it is well understood that ALD is unsurpassed as a technique for the growth of TiO2 on different structures, because it has unbeatable advantages in terms of excellent step coverage performance and uniform film growth. ALD coating is achieved by chemical reaction of hydroxyl group precursors, so that uniform films can be realized on a hydroxylated hydrophilic surface.17,18 In contrast, when a specific layer is grown on a nonhydroxylated surface, island-mode growth is revealed despite the pursuit of the layer-by-layer growth mode. Thus, to achieve thorough coating, a buffer layer with a hydroxylated surface between the two materials is essential. In this concern, n-type Al-doped ZnO thin films have been coated onto Cu2O films or NWs to act as hydrophilic interlayers in order to achieve smooth layer-by-layer growth of TiO2 and effective charge transport.4 Consequently, surface passivation of the Cu2O photocathode has been carried out by multistacking oxide films. Thus, unlike the conventional Si-, GaAs-, and InP-based semiconductors with a single TiO2 passivation layer, the interlayer plays an important role in achieving effective charge transport and transfer. Considering all the aforementioned facts, we develop a highly stable protective TiO2 surface layer on Cu2O NWs by using a strategically unique solution-based process without the use of additional layers. Remarkably robust Cu2O NWs photocathodes conformally coated with TiO2 layers on the sidewalls of intermediate Cu(OH)2 NWs are thus obtained. Experimental Section Preparation of Cu(OH)2, Cu2O NWs, and surface protective TiO2 Indium tin oxide (ITO) working electrodes with a sheet resistance of 10 Ω/sq and a thickness of 180 nm were taken as conducting substrate for depositing copper (Cu) film on it. Before depositing, these ITO/glass substrates were cut into 2 cm × 3 cm pieces and cleaned by sequential sonication in acetone, ethanol, and distilled water for 10 min each. The Cu film (1.5 μm) was grown by DC magnetron sputtering at room temperature. Further, the Cu(OH)2 NWs were prepared by anodizing Cu-coated ITO substrates in constant current density (10 mAcm−2) mode at room temperature in 3 M KOH solution for 3 min. The as-synthesized Cu(OH)2 NWs were further used to obtain Cu2O NWs in two different ways

Page 2 of 13

i.e., direct thermal oxidation (DTO) and indirect thermal oxidation (IDTO) process as shown in the schematics of Fig. 1(a). In the former process (i.e., DTO), the prepared Cu(OH)2 NWs were sufficiently converted to copper oxide by thermal annealing at 600°C under N2 gas atmosphere for 4 h, where oxidation did not induce any destruction of the nanowire arrays. Over these Cu2O NWs, a solution-processed TiO2 passivation layer was coated using the spin coating method. For this, the Ti precursor solution of 0.4 M titanium(IV) isopropoxide (C12H28O4Ti, Sigma-Aldrich) in 2methoxyethanol was prepared and magnetically stirred at 400 rpm for 12 h at 75°C. Further, it was filtered using a 0.2m syringe filter. The as-prepared solution was then completely spin-coated (three cycles) onto the Cu2O NWs at 3000 rpm for 30 s, followed by sequential soft- and hardbake sintering processes to produce condensed oxide layers. The soft-bake process of sol-precursors was carried out at 150 °C in ambient air for about five minutes and the hard baking was eventually achieved at 600oC with the ramping rate of 10 °C/min for 4 hours in an N2 atmosphere. In the latter case (i.e., IDTO), the Ti precursor solution (discussed above) was completely spin-coated (three cycles) onto the Cu(OH)2 NWs at 3000 rpm for 30 s, followed by sequential soft- and hard-bake sintering processes to produce condensed oxide layers. During the soft-bake process at 150 °C (about five minutes), the products of solvent decomposition, such as NO2, H2O, and CO2, could be evaporated. The crystalline metal-oxygen-metal frames were finally formed by transforming Cu(OH)2 to Cu2O NWs during the hard-bake process at 600 °C (ramping rate 10 oC/min) for 4 hours in an N atmosphere. The film thickness 2 of each of the TiO2 layers was confirmed to be 20 nm per cycle, and the target thickness was calibrated by monitoring the different TiO2 layers grown with varying molar concentrations on the Si substrates discussed elsewhere.19 For comparison, conventional TiO2 was grown by ALD with titanium isopropoxide and distilled water as Ti and oxygen sources, respectively. N2 was used as a purging gas. The deposition temperature was 150 °C. Characterization and measurement methods All PEC tests were performed in 1 M Na2SO4 background solution (pH 6.5) at room temperature. A solar simulator (Yamashita-Denso YSS-50A, 150 W Xenon lamp) was used as the beam source to produce an AM 1.5G spectrum, where the light intensity was calibrated to 100 mW/cm2 using a standard silicon photodetector. All PEC measurements were performed using a three-electrode system (Princeton Applied Research Versastat 4) with a Pt mesh counter electrode and an Ag/AgCl (saturated NaCl) reference electrode. Linear scan voltammetry (LSV) was performed in the photo- and dark-current (on/off pulse: 2 s) modes with a voltage sweep of 0–0.7 V vs. the reversible hydrogen electrode (RHE) at a scan rate of 10 mVs−1. The stability of the photocurrent was evaluated by measuring the photocurrent variation under illumination at a fixed electrode potential of 0 V vs. RHE. Additionally, the cyclic voltammetry (CV) sweep profiles were acquired to evaluate the variation of the PEC properties of TiO2 depending on the annealing temperature. This study helped in understanding the chemical durability as well as the photoactivity of the passivated TiO2 layer.

ACS Paragon Plus Environment

Page 3 of 13 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

The chemical bonding states of the components of the Cu(OH)2 and Cu2O NWs and TiO2 films were identified via X-ray photoelectron spectroscopy (XPS; ESCALAB 250, Thermo Fisher Scientific Co.) using an Al-Kα X-ray source. The morphological and crystalline properties of the Cu2O and Cu(OH)2 NWs were visualized by field-emission scanning electron microscopy (FE-SEM, JSM-6700F, 10 kV) and investigated by X-ray diffraction (XRD, Bruker AXS D8 Discover with a Cu-Kα radiation source), respectively. A transmission electron microscope (TEM, JEM-2100F HR) was utilized for acquisition of the nanoscale images of the Cu2O NWs. Results and Discussion Based on the Pourbaix diagram for Cu, Cu2O NW scan be synthesized via a combination of anodizing and thermal annealing.20 As illustrated in the schematic in Figure 1(a), the Cu precursor films deposited by sputtering could be chemically transformed into Cu(OH)2 NWs by reaction of the films with OH- via anodization in a basic solution. Subsequent thermal annealing under oxygen-deficient nitrogen atmosphere induced a phase change from Cu(OH)2 NWs to Cu2O NWs. This process is very simple and reproducible and has been used by Grätzel and coworkers as an electrochemical approach for the formation of Cu2O NWs. The reaction can be represented as follows: Cu + 2OH-

Anodization

Cu(OH)2 + Cu

Cu(OH)2 + 2e-

Thermal annealing

Cu2O + H2O

(1) (2)

As shown in Figure 1(b) and S1, the Cu2O crystals preferentially oriented along the (111) orientation based on the fact that the NWs were mostly vertically aligned. Peaks of residual Cu phases that did not undergo anodization and transformation were also observed in the profiles of the Cu2O NWs (Fig. S1). These Cu residues could not be completely removed via additional thermal annealing and were distributed near the bottom electrodes, as excess anodization or oxidation of Cu frequently induced detachment of the NWs from the substrate. This implies that the lower Cu residue layer plays the role of an adhesive film between Cu(OH)2 or Cu2O and the ITO electrode [Figure S2 (a)]. XPS analysis provided clear evidence of the formation of Cu(OH)2 and Cu2O phases before and after thermal annealing, respectively, as shown in Figure 1(c). The intense binding energy peaks at 932.5 and 953 eV corresponding to Cu 2p3/2 and Cu 2p1/2, respectively, indicating Cu+ for Cu2O (upper curve). In case of Cu(OH)2, the intense binding energy peaks were located at 934.5 and 954.4 eV, respectively, corresponding to Cu 2p3/2 and Cu 2p1/2states, indicating presence of Cu2+. Further, the XPS profile also indicated a weak satellite peak at 933.7 eV of Cu+ and two additional satellite peaks at 943.7 and 962.5 eV, corresponding to the paramagnetic chemical state of Cu2+ in Cu(OH)2.21 Figure 2(a) shows the LSV data (black line) for the Cu2O NWs using a photocurrent of 0.9 mAcm−2 at RHE 0 V. A moderate current of ~0.2 mAcm−2 was obtained even under dark conditions without photocarriers. This dark current was

also observed for the sample with Cu(OH)2 NWs (blue line). These currents are attributed to the electrochemical corrosion resulting from the reduction of Cu(OH)2 and Cu2O (based on the Pourbaix diagram), where the electrolytes can quickly interact with the exposed surface.22 Moreover, the electrolyte can even permeate through the Cu2O or Cu(OH)2 NWs, and an undesirable shunt path will be generated due to direct electrical contact between the bottom electrode and electrolyte, leading to degraded PEC performance. In addition, as shown in Figure 2(b), the repeated PEC measurements led to the continuous reduction of the photocurrent because of PEC corrosion of the Cu2O NWs under simultaneous photo-irradiation, applied bias, and corrosive pH conditions. The LSV data confirm that robust surface protection of the ITO/Cu/Cu2O NWs is inevitable for stable PEC operation. It is reported that Cu2O/TiO2 has suitable band-edge positions as shown in Fig. S2 (b)23-25 and importantly, TiO2 is thermally and chemically stable and can help in forming a suitable protective layer. Accordingly, as a first approach and according to earlier reports15,26-29, we deposited TiO2 thin layer (< 5 nm) using ALD technique on the Cu2O photocathodes with a nonhydroxylated surface. This was found to generate a rugged surface morphology, similar to an island growth behavior [Fig. S3 (a)]. This is due to the hydrophobicity of the surface. As a matter of fact, the PEC corrosion effect was clearly observed [Fig. S3 (b)]. Thus, the formation of ultrathin TiO2 layers (< 5 nm) is restricted.4 Hence, in order to achieve sufficient chemical passivation, the Cu2O NWs were coated with TiO2 films having a thickness >20 nm [Fig. S3 (c)]. However, it was found that the photocurrent drops [Fig. S3 (d)], which may be due to the presence of interface states that results into faster recombination effects [30]. It should be noted that this effect could also be equally found in thin layer coatings, but in the present study thin TiO2 layer coating are more corrosion prone due to breaching. Importantly, the presence of the capacitive spikes in the photocurrent indicated that the drop in photocurrent is not completely related to thickness; but is equally be related to the charging and discharging at the semiconductorelectrolyte junction due to a high charge transfer resistance. This is quite possible as there exists electron-trapping states in either the Cu2O or TiO2 that results in photocorrosive currents.4,6 Thus, this suggested that the direct coating of the Cu2O NWs with TiO2 film leads to PEC corrosion and/or reduction of the photocurrent, despite the use of chemically durable TiO2 prepared by the favorable ALD process. This indicated that if hydrophilicity is not achieved (in between photoelectrode and passivation layer), the resulting imperfect surface passivation induces a locally corrosive PEC reaction, with consequent deterioration of the PEC performance. If the issue of corrosion cannot be precluded via ALD of the passivation layer in the design of NWs-based photocathodes, the ALD process is not adequate and the development of alternative methods is necessary. Therefore, to find a suitable solution for protecting the photoelectrode surface, the first issue that we address in this study is how hydrophilic surface properties can be guaranteed to achieve thorough coating of a single TiO2 layer on Cu2O NWs, instead of forming an elaborately controlled complex multilayer structure such as Cu2O/Al-doped ZnO/TiO2.4 As alternatives to ALD-based vacuum deposition, several solution-based approaches for preparing ultra-thin oxide

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces 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

semiconductor or insulator films have been recently explored with the aim to achieve low-cost and large-area printing deposition, by using: i) high-temperature thermal sintering and ii) low-temperature ozone-free UV photoactive annealing of spin-coated precursor solutions.29-32 These approaches have been well-established for oxide thin-film transistors using transition metal-based oxide semiconductors and insulators (InZnO, InGaZnO, AlOx, TiOx, ZrOx, and HfOx). Nevertheless, to the best of our knowledge, passivation coating of nanostructured photocathodes with a sol-gel processed oxide has rarely been studied as an approach for achieving PEC cells that are stable and corrosion-free over the long-term. Of course, there are reports that suggest solution based oxide coatings; but they are utilized mostly on uniform thin film surfaces.3335Apart from this even though UV photo-activation annealing after spin-coating is very attractive for formation of a low-temperature processed passivation layer, but we preferred a thermal sintering-based sol-gel approach. This is for the reason that, UV processed oxide films might undergo rapid chemical corrosion, plausibly due to insufficient Ti-O chemical bonding and the presence of numerous impurity residues,36-39 which may not relatively be significant in thermal sintering. Thus, we selected a thermal sinteringbased sol-gel approach (solution coating + sintering at moderate temperature) as a novel solution-based strategy, which is expected to have a strong impact on the passivation of corrosive PEC electrodes as it is a simple and vacuumfree low-cost process. It is anticipated that precursor solutions with low viscosity can readily permeate into the spaces between the NWs and can completely permeate the wires if the surface can be imbued with hydrophilic character for solution coating, regardless of the morphology and length of the NWs. However, regarding solution-based coating, the following must be considered: 1) preferential chemical corrosion of incomplete bonding states as a result of an inadequate sintering process, and 2) how to sufficiently wet the side walls. Firstly, to attain durable TiO2 thin films via sol-gel coating, the sintering process must be optimized because solution-processed oxide films undergo fast PEC corrosion and exhibit low-electrical performance due to the incomplete formation of M-O bonds, ligand-induced impurities, and low film density. Hence, sol-gel processed TiOx with wellbonded Ti-O-Ti networks is required to prevent corrosion of the PEC and achieve effective carrier transfer between the photocathode and electrolyte. The activation energy-driven chemical reaction history of the precursor solution determines the chemical bonding state of sol-gel processed oxide films. The method of determining the sintering temperature for densification reaction is explained in the Supplementary Information [see Figure S4 (a)-(c)]. In order to have a proper control over the TiO2 coating together with its chemical durability and photoactivity, we performed cyclic voltammograms (CV) studies for the ~40 nm thick TiO2 films sol-gel-prepared on the ITO/glass substrates with various sintering temperatures. In addition to this a repetition test was also carried out with ten cycles per sample under the same voltage sweep and illumination conditions, as shown in Figure S5 [(a) and (b)]. Because conventional TiO2 layers are n-type semiconductors and are photoactive, the photoanodic current must be determined from common TiO2 films. However, as shown in Figure S5 (b), annealing at

Page 4 of 13

temperatures up to 400 °C led to no meaningful photoanodic current in the positive bias region and the CV profiles were quite similar to that obtained with the individual ITO electrode, resulting in degradation and corrosion of the TiO2 films. This is equally demonstrated by the XPS results [Figure S4 (b) and (c)], densification and grain growth, which occurred above 400 °C, led to saturated/strong Ti-O bonding and better oxidation. Here, the TiO2 samples annealed at 500 and 600 °C exhibited distinct anodic currents and reproducible CV curves, and thus a sintering temperature of 600 °C was selected for coating the TiO2 films onto the Cu2O NWs. It was revealed that the solutionprocessed TiO2 and annealed at 600 oC were polycrystalline (anatase crystal phase) with rough morphology and randomly-oriented grains as is shown in the TEM and SAED patterns [Fig. S4 (d)]. After having proper analysis and understanding of the solution-processed TiO2, we prepared the first sample by the typical process flow: absorber: Cu2O nanowire formation (DTO)→surface passivation: TiO2 sol-gel coating and sintering (as mentioned in experimental section). In this case, the Cu2O NWs and the lower Cu residue layer could be simultaneously covered with sol-gel coated TiO2; thus, this process was expected to effectively suppress generation of the dark current and shunt path [Figure 3 (a)], and the initial photocurrent was also improved (2.8 mAcm−2 vs. RHE 0 V). Nevertheless, with repeated LSV measurements, a considerable drop in the photocurrent was observed, indicating the generation of photo-corrosive reaction sites in a short period. The stability test indicated that for the TiO2 sol-gel coated directly on Cu2O NWs, corrosive sites where the Cu2O and electrolyte make contact (Figure S5) were still present despite the use of the chemically-robust sol-gel processed TiO2 films. Fortunately, the absence of the dark current suggests that the bottom Cu residue layer was fully covered by sol-gel TiO2 and the corrosive sites probably originate from the side-walls of the NWs. As propounded previously, the issues of an incomplete conformal coating of the side walls of the NWs must be addressed, considering the role of the interlayers (such as n-type Al-doped ZnO) in the ALD TiO2 film. Thus, the wettability of the Cu2O nanowire arrays by the Ti[OCH(CH3)2]4 precursor solution was evaluated via contact angle measurements, as shown in the inset of Figure 3 (a). The Cu2O NWs with Cu-O surficial bonds exhibited low wettability by the TiO2 precursor solution, as indicated by the high contact angle. This may result in partial exposure of the side walls of the Cu2O NWs due to non-complete coating by the TiO2 precursor. This is attributed to the intrinsic differences between Cu2O and TiO2, and thus the underlying surface for ALD of the TiO2 passivation layer must be endowed with functionally hydrophilic characteristics. To overcome this issue and to avoid any additional film coating such as Al-doped ZnO, we employed a versatile approach by using hydrophilic alcohol-based TiO2 precursor solutions for coating the hydrophilic surface by minimizing the surface energy via the formation of intermediate Cu hydroxide compounds. Before the formation of the Cu2O NWs, Cu(OH)2 NWs were preferentially produced as an intermediate phase (prior to transformation), as shown in Figure 1(a). Such synthesized Cu(OH)2 NWs could not produce a marginal photocurrent (blue line) as it is inherently photo-inactive and insulating, as shown in Figure

ACS Paragon Plus Environment

Page 5 of 13 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

2(a). However, on the other hand, the Cu(OH)2 intermediate phase possess hydroxyl groups that suitably confer the desired hydrophilicity to the surface, unlike the case of nonhydroxylated Cu2O. To understand this, the wettability of the Cu(OH)2 NWs by the Ti precursor solution was studied and it showed remarkable improvement, as indicated by the low contact angle [see inset of Figure 3 (b)], due to the hydrophilic Cu-based functional groups (Cu-OH). Thus, these Cu(OH)2 NWs may be able to promote conformal solgel coating of the TiO2 precursor solutions without exposed sites. With this approach, we could easily manage to thoroughly coat Cu(OH)2 NWs with TiO2 by simple solutionprocess and thermally treatment as mentioned in experimental section [Fig. 1 (a)]. Further, the PEC study of such prepared Cu2O NWs was carried out. Interestingly, Figure 3(b) reveals almost identical curves and the absence of a dark current in the repeated LSV tests. Moreover, it exhibited a higher onset potential of 0.55 V vs. RHE and a photocurrent of ~1.5 mAcm−2 at 0 V vs. RHE. TiO2 passivation revealed higher onset potential indicating enhanced carrier blocking ability and reduced electron back transfer.40,41 It should be noted that sintering of the Cu(OH)2 nanowire/sol-gel TiO2 bilayers led to excellent and stable PEC performance of the resulting oxide system without any degradation of the photocurrent during the LSV measurements. The SEM and TEM images shown in Figure 4 [(a)-(b)] reveal the formation of distinguishable Cu2O/TiO2 interfaces. This observation indicates that a sufficient and stable supply of oxygen was maintained during the thermal process, which induced the phase change from Cu(OH)2 to Cu2O, and the substantial photocurrent of the final product confirms complete phase transformation to Cu2O. It should be noted that since the initial NWs [Fig. 1 (b)] have non-uniform and rough side-walls, the coating of TiO2 over it conformed a similar growth pattern [Fig. 4 (a)-(b)]. However, it was ensured that a thorough TiO2 coating on Cu2O was achieved [Fig. 4 (b)]. Consequently, the Cu2O/TiO2 NWs made from hydrophilic Cu(OH)2 displayed excellent photo-chemical stability during testing for more than one day (33 h), as shown in Figure 4(c). In contrast, the samples prepared by direct passivation on the Cu2O NWs resulted in a rapid photocurrent drop to