Nanoporous 6H-SiC Photoanodes with a Conformal Coating of Ni-FeOOH Nanorods for Zero-Onset-Potential Water Splitting.

A surface-nanostructured semiconductor photoelectrode is highly desirable for photoelectrochemical (PEC) solar-to-fuel production due to its large active surface area, efficient light absorption, and significantly reduced distance for charge transport. Here, we demonstrate a facile approach to fabricate a nanoporous 6H-silicon carbide (6H-SiC) photoanode with a conformal coating of Ni-FeOOH nanorods as a water oxidation cocatalyst. Such a nanoporous photoanode shows significantly enhanced photocurrent density (jph) with a zero-onset potential. A dendritic porous 6H-SiC with densely arranged holes with a size of ∼40 nm on the surface is fabricated by an anodization method, followed by the hydrothermal deposition of FeOOH nanorods and electrodeposition of NiOOH. Under an illumination of AM1.5G 100 mW/cm2, the Ni-FeOOH-coated nanoporous 6H-SiC photoanode exhibits an onset potential of 0 V versus the reversible hydrogen electrode (VRHE) and a high jph of 0.684 mA/cm2 at 1 VRHE, which is 342 times higher than that of the Ni-FeOOH-coated planar 6H-SiC photoanode. Moreover, the nanoporous photoanode shows a maximum applied-bias-photon-to-current efficiency (ABPE) of 0.58% at a very low bias of 0.36 VRHE, distinctly outperforming the planar counterpart. The impedance measurements demonstrate that the nanoporous photoanode possesses a significantly reduced charge-transfer resistance, which explains the dramatically enhanced PEC water-splitting performance. The reported approach here can be widely used to fabricate other nanoporous semiconductors for solar energy conversion.

Introduction

Photoelectrochemical (PEC) water splitting is regarded as a promising approach to convert solar light into a stable, environmentally friendly, renewable, and clean chemical energy, H2.[1−4] Since the pioneering PEC work by Fujishima and Honda,[5] semiconductor materials such as TiO2, Fe2O3, ZnO, and BiVO4 have been widely studied for PEC water splitting.[6−21] However, the inherent drawbacks of the semiconductor materials including the unfavorable energy band position with respect to water redox potentials, the short carrier diffusion length, and the sluggish oxygen-evolution kinetics hamper the significant progress of PEC water splitting.[22]

Silicon carbide (SiC) is a promising photoelectrode material for PEC water splitting because its energy band positions ideally straddle the water redox potentials.[23−27] Recently, commercially available 6H-SiC material has attracted considerable attention in solar-driven water splitting. In 1997, Lauermann et al. reported that both p- and n-type 6H-SiC photoelectrodes were able to support the PEC water-splitting reaction.[28] Later on, Akikusa et al. showed a PEC system with a p-type 6H-SiC photocathode and an n-type TiO2 photoanode, which exhibits a maximum jph of 0.05 mA/cm2 and a solar-to-hydrogen (STH) conversion efficiency of 0.06%.[29] Kato et al. showed a low STH efficiency of 0.001% for a 6H-SiC photoelectrode connected to a Pt electrode.[30] So far, the reported 6H-SiC photoelectrodes show a rather low STH conversion efficiency (Table S1). Generally, the indirect band gap of 6H-SiC has a smaller light absorption coefficient than direct band gap semiconductors. This results in larger light penetration depth than the sum of the width of the space-charge region and the carrier diffusion length. As a result, most photogenerated carriers cannot be separated and collected for PEC water splitting. Moreover, the sluggish water oxidation reaction on the photoanodes has been regarded as a bottleneck for PEC reactions. In this work, we simultaneously address these issues by the fabrication of nanoporous 6H-SiC photoanodes with a coating of highly active water oxidation cocatalyst Ni–FeOOH nanorods.

Porous semiconductors are commonly fabricated by the anodization method. In the early 1990s, Shor et al. studied the fabrication of nanoporous SiC by anodization.[31−35] Later on, Choyke et al. reported a hybrid columnar and dendritic porous structure of n-type 6H-SiC by controlling voltage and current density.[36] So far, much effort has been devoted to optimizing the pore structures for their optoelectronic applications.[37] However, PEC water splitting using porous SiC has not been studied.

Compared to the planar 6H-SiC (SC), the porous 6H-SiC (PSC) is expected to have an enhanced PEC performance owing to its low reflectivity on the surface, efficient light absorption as well as the dramatic increase of the surface area. Moreover, the cocatalyst coated on the porous 6H-SiC further improves the water-splitting efficiency. Recently, many earth-abundant transition-metal (Fe, Co, Ni) oxides and hydroxides have been developed as remarkable water oxidation cocatalysts.[38]

In this work, we employ a facile anodization approach to fabricate PSC photoanodes with an optimal depth of the nanoporous layer that matches the light penetration depth. Then, Ni–FeOOH nanorods with a diameter of 20–30 nm are deposited on PSC photoanodes as the water oxidation cocatalyst. Under an illumination of AM1.5G 100 mW/cm2, the Ni–FeOOH-decorated nanoporous photoanode exhibits 342 times higher photocurrent than the Ni–FeOOH-decorated planar photoanode at 1.0 V versus the reversible hydrogen electrode (VRHE). Given the fact that the anodization method can be used to fabricate a porous structure on a variety of semiconductors, this work demonstrates a simple approach to improve the water-splitting efficiency by rational design of the porous photoelectrodes.

Experimental Section

Pretreatment of 6H-SiC

Commercial single-crystal n-type 6H-SiC (0001) wafers (SiCrystal) with a thickness of ∼350 μm and a nitrogen doping concentration of 3 × 1017 cm–3 were cut into small pieces (0.25 cm2). Then, these samples were successively cleaned by acetone, ethanol, RCA1 (H2O/H2O2/NH3 = 5:1:1), and RCA2 (H2O/H2O2/HCl = 6:1:1) for 5 min each. Finally, the samples were put into 5% HF for 1 min to remove surface-native oxides and then dried by pure N2.

Preparation of 6H-SiC Electrodes

Ohmic contacts were made on the backside (C-face) of 6H-SiC by deposition of a 60-nm Ni/150-nm Au film using a vacuum evaporator, followed by annealing at 900 °C for 10 min in a N2 atmosphere. Figure S1 shows the linear current–voltage curve, confirming the formation of the good ohmic contacts on the backside of 6H-SiC. Figure S2 illustrates the fabrication process of the 6H-SiC photoanodes. The backside ohmic contact was connected to Cu tape and the whole sample was fixed on a plastic plate. The exposed edges of the photoelectrodes were sealed with epoxy resin so that only the front surface (Si-face) of 6H-SiC was exposed to an electrolyte and light for PEC measurements.

Fabrication of Nanoporous 6H-SiC by Anodization

Nanoporous 6H-SiC samples were fabricated by the anodization process in a two-electrode cell in a 5% HF solution, where the Si-face 6H-SiC (0001) was used as the photoanode, a 1 × 1 cm2 platinum plate as the cathode, and a 410-nm light-emitting diode (LED) (400 mW/cm2) as the light source. The voltage between the two electrodes is provided using a power source (BK precision 9184). To control the structure and morphology of porous 6H-SiC, the supplied voltage was changed from 6 to 24 V with a step of 6 V for a duration of 10 min. The fabricated nanoporous 6H-SiC samples, denoted as PSC6, PSC12, PSC18, and PSC24, represent the samples etched at 6, 12, 18, and 24 V, respectively. Also, the pristine planar 6H-SiC (denoted as SC) was used as a reference. After PEC etching, samples were cleaned by deionized water and dried by blowing nitrogen.

Deposition of Ni–FeOOH Cocatalyst on 6H-SiC Photoanodes

First, FeOOH was deposited on both planar and porous 6H-SiC by a hydrothermal method. For the deposition of FeOOH, 25 mL of precursor solution was prepared with FeCl3 (30 mM) and urea (45 mM). The 6H-SiC photoanodes were immersed in the solution and the hydrothermal reactor was kept at 100 °C in an oven for 1 h. Then, the FeOOH-coated 6H-SiC photoanodes were rinsed by deionized water and dried by blowing pure N2. Second, the NiOOH layer was deposited on the FeOOH/6H-SiC photoanodes by a photo-assisted electrodeposition process in a 0.1 mol/L NiCl2 solution (pH = 7) at 0.4 V versus Ag/AgCl for one hour under AM1.5G 100 mW/cm2 illumination, where an FeOOH/6H-SiC photoanode, a 1 × 1 cm2 platinum plate, and Ag/AgCl (saturated KCl) were used as the working electrode, the counter electrode, and the reference electrode, respectively. Finally, the Ni–FeOOH-coated 6H-SiC photoanodes were rinsed by deionized water and dried by blowing pure N2.

Characterizations

UV–vis absorption spectra were measured by a PerkinElmer Lambda 950 UV/VIS setup. Scanning electron microscopy (SEM) and energy-dispersive spectroscopy (EDS) images were measured using a LEO 1550 Gemini instrument, with a voltage of 10 kV, a WD of 8.5 mm, and an X-Max silicon drift detector (Oxford Instruments). High-resolution transmission electron microscopy (HRTEM) was carried out using a JEM 2100F instrument with a voltage of 200 kV. X-ray diffraction (XRD) patterns were measured by a Prefix diffractometer with Cu Kα1 (λ = 1.54 Å). X-Ray photoelectron spectroscopy (XPS) was measured by an ESCALab220I-XL spectrophotometer with Al-Kα radiation.

Photoelectrochemical Measurements

Photoelectrochemical (PEC) measurements were performed in a 1 M NaOH (pH ≈ 13.6) solution under AM1.5G 100 mW/cm2 illumination in a three-electrode cell, where the photoanode was used as a working electrode, a 1 × 1 cm2 platinum plate as a counter electrode, and a standard Ag/AgCl (saturated KCl) electrode as the reference electrode. The solution was purged with high-purity (99.999%) Ar gas for over 30 min prior to PEC measurements to remove the dissolved oxygen. For the PEC measurements, the electrochemical workstation (Princeton Applied Research VersaSTAT 3) was used and the light of AM1.5G 100 mW/cm2 was generated from the solar simulator (LOT-Quantum Design GmbH), which was calibrated using a standard Si photovoltaic cell. The current density–voltage (j–V) curves were measured with a scan rate of 30 mV/s. The potential measured versus the Ag/AgCl reference electrode (VAg/AgCl) was converted into the potential versus reversible hydrogen electrode (VRHE) using the following equation:The current density–time (j–t) curves were measured under AM 1.5G 100 mW/cm2 illumination at 1 VRHE. The O2 and H2 gases during the PEC water splitting were measured by a microgas chromatograph (Agilent Technologies Micro GC490) using Ar as the carrier gas.

Results and Discussion

Figure compares the top-view and cross-sectional SEM images of the planar 6H-SiC (SC) and the nanoporous 6H-SiC samples (PSC6, PSC12, PSC18, PSC24), which were prepared by the anodization method at a constant voltage of 6, 12, 18, and 24 V, respectively. As can be seen in Figure a–d, on increasing the voltage to 18 V, densely arranged holes with a diameter of ∼40 nm appear on the surface of 6H-SiC (Figure c–e). Further increase of the applied voltage to 24 V results in a slight increase of the hole diameter to ∼50 nm (Figures e and S3). The cross-sectional SEM images in Figure g–j show that the depth of the porous layer in PSC6, PSC12, PSC18, and PSC24 is 1.85, 5.68, 12.40, and 30.98 μm, respectively. This reveals that the depth of the porous layer is gradually increasing with an increase in the applied voltage.

Figure 1

SEM images of the planar 6H-SiC (SC) and nanoporous 6H-SiC samples (PSC6, PSC12, PSC18, PSC24) prepared at 6, 12, 18, and 24 V, respectively. Top-view SEM images of SC (a), PSC6 (b), PSC12 (c), PSC18 (d), PSC24, (e) and cross-sectional SEM images of SC (f), PSC6 (g), PSC12 (h), PSC18 (i), and PSC24 (j). High-magnification images (k–o) measured on the denoted squares of images (f–j).

As can be seen in Figure g–j, PSC6 shows a horizontally etched pattern, while PSC12, PSC18, and PSC24 samples exhibit a dendritic structure. Choyke et al. reported that the preferential etching direction of Si-face 6H-SiC is [112̅0] (the horizontal direction), which has a higher etching rate than that in the [0001̅] direction (vertical).[39] This explains why we first observe the formation of horizontal porous patterns (Figure l). With further etching, the pores branch and grow, thus forming the dendritic structure due to the faster horizontal etching than vertical etching (Figure m–o). To confirm this etching mechanism, we studied the effect of etching time on the formation of pore structures. Figure S4 compares the cross-sectional SEM images of the nanoporous 6H-SiC etched at 12 V for 1 and 10 min, respectively. The sample etched for 1 min clearly shows the same horizontal structure as the sample PSC6, while the prolongation of the etching time to 10 min results in the formation of the dendritic pattern. This result reveals that the dendritic porous structure is formed due to the faster horizontal etching rate than the vertical one.

To demonstrate the effect of the porous structure in enhancing the PEC water-splitting performance, we chose the PSC18 sample to compare with the planar sample SC. The depth of the porous layer in PSC18 is 12.40 μm, which is an optimal value for the sunlight penetration depth. Due to a band gap of 3.02 eV, 6H-SiC mostly absorbs ultraviolet light and the light penetration depth is smaller than ∼12 μm at wavelengths shorter than 360 nm (3.44 eV).[40] This indicates that most light with photon energies higher than the band gap of 6H-SiC will be absorbed within a depth of 12 μm. The additional thicker porous layer (>12 μm) does not contribute to light harvesting but only increases carrier recombination in the porous structure. Moreover, the large etched holes observed in PSC24 (Figure o) indicate the deterioration of the porous structure.

To passivate the surface defects and increase the water oxidation activity, earth-abundant Ni–FeOOH was used as a water oxidation cocatalyst to deposit on both the planar 6H-SiC (Ni–FeOOH/SC) and the nanoporous PSC18 sample (Ni–FeOOH/PSC18) under the same condition. Figure shows the top-view and cross-sectional SEM images of Ni–FeOOH/SC and Ni–FeOOH/PSC18. Both exhibit a layer of Ni–FeOOH nanorods. On the planar SC surface, the Ni–FeOOH nanorods show a diameter of 30–40 nm with a layer thickness of 330 nm. On the porous sample PSC18, we observe relatively uniform and small Ni–FeOOH nanorods with a diameter of 20–30 nm. Considering that the porous sample PSC18 exhibits holes with a diameter of 40 nm on the surface, the Ni–FeOOH nanorods might start to grow in the pore structure and thus are constrained by those holes. Therefore, the diameter of the FeOOH nanorods grown on the porous surface is smaller and relatively uniform than that grown on the planar surface. Moreover, the same growth condition gives rise to much thinner Ni–FeOOH nanorods (∼120 nm) on the porous surface than that on the planar surface, further indicating that the Ni–FeOOH nanorods might grow in the dendritic porous structure. To confirm this, EDS mapping was employed to characterize the Ni–FeOOH nanorods on the SC and PSC18 samples. Figure S5 compares the EDS mapping of C, Fe, O, and Ni elements for cross-section images of Ni–FeOOH/SC and Ni–FeOOH/PSC18. It shows the existence of Fe, O, and Ni elements only on the surface of the planar SC sample. In contrast, the porous sample PSC18 exhibits the existence of Fe, O, and Ni elements on both the surface and inside the nanoporous structure. This result confirms that Ni–FeOOH also grows in the dendritic porous structure.

Figure 2

SEM images of the Ni–FeOOH-coated planar 6H-SiC (Ni–FeOOH/SC) and the Ni–FeOOH-coated nanoporous sample PSC18 (Ni–FeOOH/PSC18). Top-view image (a), cross-sectional image (b), and histogram of the Ni–FeOOH diameter distribution, and (c) of the Ni–FeOOH/SC. Top-view image (d), cross-sectional image (e), and histogram of the Ni–FeOOH diameter distribution, and (f) of Ni–FeOOH/PSC18.

The structure and crystallography of the as-prepared Ni–FeOOH nanorods were characterized by HRTEM. Figure shows the Ni–FeOOH nanorods with a diameter of 40 nm, consistent with the SEM results. The HRTEM image reveals a clear lattice spacing of 0.74 nm of the nanorod that corresponds to the (110) plane of the β-FeOOH crystallography phase.[41] We also observe that a part of the nanorods shows no clear lattice fringes, which might indicate the formation of amorphous NiOOH.

Figure 3

High-resolution TEM image of the Ni–FeOOH nanorods.

The oxidation state and chemical composition of the Ni–FeOOH layer were studied by XPS measurements. The XPS survey spectra in Figure a confirmed the existence of Fe, Ni, and O elements, consistent with the EDS results. As can be seen in Figure b, the spectrum of Fe 2p shows two major peaks (711 eV for Fe 2p3/2 and 725 eV for Fe 2p1/2) with their shake-up satellites (719 and 733 eV). These peaks agree very well with the reported characteristics peaks of Fe3+ in FeOOH.[42] The Ni 2p spectrum shown in Figure c is dominated by the 2p3/2 peak at 856 eV and its shake-up satellite peak around 861 eV, which are consistent with the Ni3+ peaks in NiOOH.[43] The O 1s XPS spectrum (Figure d) can be fitted by two peaks corresponding to O2– (530.3 eV) and OH– (531.6 eV).[26,44,45] The similar area of OH– and O2– peaks conforms to the theoretical ratio of 1:1 between OH– and O2– in Ni–FeOOH. This result also indicates the formation of Ni–FeOOH. Therefore, together with the SEM and TEM results discussed above, the XPS results confirm the formation of Ni–FeOOH nanorods on 6H-SiC.

Figure 4

XPS spectra of Ni–FeOOH deposited on 6H-SiC. XPS full survey spectrum (a), Fe 2p (b), Ni 2p (c), and O 1s (d) spectra of Ni–FeOOH.

Figure compares the PEC performance of the planar Ni–FeOOH/SC and porous Ni–FeOOH/PSC18 photoanodes. Under an illumination of AM1.5G (100 mW/cm2), the Ni–FeOOH/PSC18 photoanode exhibits a photocurrent density of 0.684 mA/cm2 at 1 VRHE, which is 342 times higher than the photocurrent density (0.002 mA/cm2 at 1 VRHE) of the Ni–FeOOH/SC photoanode. Moreover, for both planar and nanoporous photoanodes, the onset potentials of the Ni–FeOOH-deposited photoanodes are negatively shifted to ∼0 VRHE compared to the onset potentials of ∼0.2 VRHE for photoanodes without the Ni–FeOOH cocatalyst (Figure S6). This result indicates that the Ni–FeOOH cocatalyst efficiently reduces the overpotential for water oxidation. The Mott–Schottky plots of Ni–FeOOH/SC and Ni–FeOOH/PSC18 photoanodes show that the flat-band potential is around −0.22 to −0.38 VRHE (Figure S7), which is very close to the reported flat-band potential of −0.32 VRHE for 6H-SiC.[46]

Figure 5

PEC performance of the Ni–FeOOH-coated planar 6H-SiC (Ni–FeOOH/SC) and Ni–FeOOH-coated PSC18 (Ni–FeOOH/PSC18) photoanodes. Chopped current density–voltage (j–V) curves of the Ni–FeOOH/SC (a) and Ni–FeOOH/PSC18 (b) photoanodes. Note the unit of μA/cm2 for the photocurrent in (a) and the unit of “mA/cm2” for (b). Applied-bias-photon-to-current efficiency (ABPE) plots of Ni–FeOOH/SC (c) and Ni–FeOOH/PSC18 (d). All of the PEC measurements are done in a 1.0 M NaOH solution under AM1.5G 100 mW/cm2 illumination.

Figure c–d compares the ABPE plots of the Ni–FeOOH/SC and Ni–FeOOH/PSC18 photoanodes. The ABPE is given by the equation: ABPE = jph × (1.23 – |V|)/PAM1.5G, where jph, V, and PAM1.5G stand for the photocurrent density, applied potential, and light density of AM1.5G illumination, respectively. As can be seen in Figure d, the Ni–FeOOH/PSC18 photoanode exhibits a maximum ABPE of 0.582% at 0.36 VRHE, which is 342 times higher than the maximum ABPE (0.0017%) of the Ni–FeOOH/SC photoanode. These results clearly demonstrate that the Ni–FeOOH coating on the porous 6H-SiC photoanode significantly improves the overall PEC water-splitting performance. The generated H2 and O2 volumes of the Ni–FeOOH/PSC18 photoanode were measured at 1 VRHE under a steady-state illumination of AM1.5G 100 mW/cm2 in a 1.0 M NaOH solution (Figure ). The Faradaic efficiencies for H2 and O2 evolutions are ∼98 and ∼70%, respectively. The photocurrent of the planar Ni–FeOOH/SC photoanode was too low to measure H2 and O2 evolutions by gas chromatography.

Figure 6

PEC water splitting by the Ni–FeOOH/PSC18 photoanode. (a) Current density versus time (j–t) curve of the Ni–FeOOH/PSC18 photoanode at 1 VRHE under steady-state AM1.5G 100 mW/cm2 illumination in a 1.0 M NaOH solution. (b) Measured H2 and O2 volumes during (a). The dotted lines show the theoretical volumes of H2 and O2 with 100% faradaic efficiencies, respectively.

Electrochemical impedance spectroscopy (EIS) measurements were employed to study the charge-transfer properties at the interface of the photoanode and electrolyte. The EIS measurements were carried out at 0.4 VRHE in the frequency range of 10–105 Hz. Figure shows the Nyquist plots of Ni–FeOOH/SC and Ni–FeOOH/PSC18 photoanodes under AM1.5G 100 mW/cm2 illumination. In the Nyquist plots, semicircles show the charge-transfer properties from the photoanode to the electrolyte. The smaller diameters of the semicircles indicate smaller charge-transfer resistances. As can be seen in Figure , the EIS data are fitted by an equivalent circuit shown in the inset of Figure b, which consists of the series resistance (Rs), the charge-transfer resistance (Rct) from the semiconductor to the surface, the constant-phase-element of the space-charge capacitance (CPESC), the charge-transfer resistance from the photoanode to the electrolyte (Rct, trap), and the corresponding capacitance (CPEtrap). The fitted values of Rct and Rct, trap for the Ni–FeOOH/PSC18 photoanodes are significantly decreased to 48 and 231 Ω cm2, which are much smaller than those values (Rct = 1000 Ω cm2 and Rct, trap = 6 80 000 Ω cm2) for the planar Ni–FeOOH/SC photoanode, as can be seen in Table S2. The EIS results clearly demonstrate that the porous Ni–FeOOH/PSC18 photoanode possesses a significantly reduced charge-transfer resistance, which explains the dramatically enhanced PEC water-splitting performance.

Figure 7

Nyquist plots for the Ni–FeOOH/SC (a) and Ni–FeOOH/PSC18 (b) photoanodes measured at 0.4 VRHE under AM1.5G 100 mW/cm2 illumination.

To better understand the enhanced PEC water-splitting performance of the porous photoanode, we schematically compare the light absorption and carrier transport properties in the planar and porous photoanodes (Figure ). In the planar photoanode, the direction of light absorption is the same as that of charge-carrier collection (Figure a). To be collected for PEC water splitting, the photogenerated carriers must be separated by the built-in electric field and be swept toward the surface or backside before recombination. Due to the indirect band gap of 6H-SiC, the light penetration depth (Dλ) is usually larger than the sum of the width of the depletion region (Wdep) and carrier diffusion length (LD). Thus, most photogenerated carriers cannot diffuse into the space-charge region to be separated and harvested for the PEC water splitting. In contrast, the porous structured photoanode decouples the direction of the carrier diffusion from the light penetration direction. As can be seen in Figure b, the porous structure significantly reduces the distance that minority carriers must travel, thus enhancing the carrier separation and collection efficiency despite the large light penetration depth and short carrier diffusion lengths. Also, the dendritic porous structure of 6H-SiC can reduce the light reflection and provide an additional light absorption path for reflected photons within pores. Figure S8 shows that the porous 6H-SiC significantly increases the light absorption compared to the planar sample.

Figure 8

Schematic illustrations of the planar Ni–FeOOH/SC and porous Ni–FeOOH/PSC18 photoanodes for PEC water splitting. The planar photoanode exhibits a high light reflection at the surface, while the porous photoanode traps light within the pore structure. The penetration depth (Dλ) of light, the width of the space-charge region (Wdep), and the carrier diffusion length (LD) are present in planar and porous photoanodes.

The incident photon-to-current efficiency (IPCE) was measured at 1 VRHE to evaluate the PEC water-splitting performance of the nanoporous photoanode (Figure S9). Under the illumination of 1 mW/cm2 monochromatic light, Ni–FeOOH/PSC18 exhibits IPCE values of 25% and 12% at 410 and 450 nm monochromatic light, respectively. However, the Ni–FeOOH/SC photoanode displays too low photocurrent to detect the IPCE spectra in the same condition. This result evidences that the nanoporous SiC photoanode exhibits improved light absorption and charge separation/collection efficiency, thus enhancing the photon-to-current conversion efficiency.

Moreover, the dendritic porous structure with a large surface area could increase the photoanode/electrolyte interface area for water splitting. To quantify the effective electrochemically active surface area of the Ni–FeOOH/SC and Ni–FeOOH/PSC18 photoanodes, we measured the capacitive current of the photoanodes versus scan rates to extract the double-layer capacitance (Cdl) according to the method reported in pieces of literature.[47,48] The increase of the surface area of the nanoporous photoanode can be estimated by comparing the values of Cdl. As can be seen in Figure S10, the Cdl values of Ni–FeOOH/SC and Ni–FeOOH/PSC18 are 0.53 ± 0.02 and 90.2 ± 9.3 μF/cm2, respectively. This comparison reveals that the electrochemically active surface area of Ni–FeOOH/PSC18 is 170 times larger than that of Ni–FeOOH/SC. This result further highlights the vital role of the nanoporous photoanode in improving the overall PEC water-splitting performance.

Conclusions

In summary, we have demonstrated a facile approach to fabricate a nanoporous 6H-SiC photoanode with a conformal coating of Ni–FeOOH nanorods as the cocatalyst, which shows a significantly enhanced PEC water splitting with the zero-onset potential. Nanoporous 6H-SiC with densely arranged holes with a size of ∼40 nm on the surface can be fabricated by an anodization method. The cross-sectional SEM results reveal that the porous 6H-SiC has a dendritic structure, formed due to the slower etching rate in the [0001̅] direction than that in the [112̅0] direction. The SEM, TEM, and XPS results confirm the deposition of Ni–FeOOH nanorods not only on the 6H-SiC surface but also in the dendritic porous structure. Under an illumination of AM1.5G 100 mW/cm2, the Ni–FeOOH-coated porous 6H-SiC photoanode exhibits an onset potential of 0 VRHE and a high photocurrent density of 0.684 mA/cm2 at 1 VRHE, which is 342 times higher than that of the Ni–FeOOH-coated planar photoanode (0.002 mA/cm2 at 1 VRHE). Moreover, a maximum ABPE of 0.582% has been achieved for the porous Ni–FeOOH/PSC18 photoanode at a rather low bias of 0.36 V, distinctly outperforming the planar counterparts. By measuring light absorption spectra and the electrochemically active surface area, we find that the nanoporous photoanode exhibits improved light absorption and 170 times larger active surface area than the planar photoanode. The EIS results clearly demonstrate that the porous Ni–FeOOH/PSC18 photoanode possesses a significantly reduced charge-transfer resistance, which explains the dramatically enhanced PEC water-splitting performance. This research highlights the vital role of the porous photoanode in improving the overall PEC water-splitting performance and provides new insights toward efficient solar-fuel generation.