Reliable bi-functional nickel-phosphate /TiO2 integration enables stable n-GaAs photoanode for water oxidation under alkaline condition

Hydrogen is one of the most widely used essential chemicals worldwide, and it is also employed in the production of many other chemicals, especially carbon-free energy fuels produced via photoelectrochemical (PEC) water splitting. At present, gallium arsenide represents the most efficient photoanode material for PEC water oxidation, but it is known to either be anodically photocorroded or photopassivated by native metal oxides in the competitive reaction, limiting efficiency and stability. Here, we report chemically etched GaAs that is decorated with thin titanium dioxide (~30 nm-thick, crystalline) surface passivation layer along with nickel-phosphate (Ni-Pi) cocatalyst as a surface hole-sink layer. The integration of Ni-Pi bifunctional co-catalyst results in a highly efficient GaAs electrode with a ~ 100 mV cathodic shift of the onset potential. In this work, the electrode also has enhanced photostability under 110 h testing for PEC water oxidation at a steady current density Jph > 25 mA·cm−2. The Et-GaAs/TiO2/Ni-Pi║Ni-Pi tandem configuration results in the best unassisted bias-free water splitting device with the highest Jph (~7.6 mA·cm−2) and a stable solar-to-hydrogen conversion efficiency of 9.5%.

eV can result from arsenic bonded to gallium (Ga-As), metallic arsenic (As 0 ), and small amount of surface native oxides (As2O3 and As2O5), respectively. 1For the Et-GaAs film, the peaks at 41.5 eV and 45 eV can be deconvoluted into four key components, including As-Ga, As 0 , and O-As-O (As2O3(As +3 ) and As2O5(As +5 )) bonding. 2 It is further calculated quantitatively by the integral area of the deconvoluted peaks.The area of As2O3(+3) is higher than As2O5(+5) in both the bare and ET-GaAs films.The calculated portion of each oxide is [5%(+3), 35%(+3)] and [1.8%(+5), 27%(+5)] for As2O3 and As2O5, respectively, in both photoanode films.Thus, the greater increase of the arsenic oxides at the Et-GaAs film is clearly evident during electrochemical etching.In particular, the position of each component being the same as in the pristine GaAs film demonstrates that the electrochemical etching process has no influence on the intrinsic nature of the GaAs, whereas the surface morphology of the films was largely modified.Further, the Ga3d spectra (Supplementary Fig. 3(c)) of the GaAs and Et-GaAs films are fitted with multiple peaks, and the peak of Ga3d5/2 at binding energy 19.17 eV comes from the Ga metal in GaAs.Also, the shoulder at the binding energy of 20.98 eV corresponds to native Ga-O in the GaAs surface. 3anwhile, in the case of Et-GaAs film, the same binding energy at 19.17 eV was observed from Ga3d5/2, but the significant variation from the shoulder peak was observed in the binding energy of 20.29 eV, due to the native gallium oxide (Ga-O).Next, in the case of the O1s spectra (Supplementary Fig. 3(d  The etched samples show the positive shift of onset potential, compared to that of bare GaAs film, closely related to the initial surface kinetic barrier, due to the formation of native oxide during the etching process.In addition, the LSV curve has the same shape for the Et-GaAs-10min and 20 min, in which the difference of Jph can be explained by the depth of surface etching sufficiently affecting the surface features such as porosity, roughness and surface area etc. Conversely, the complete opposite trend is remarkably noticed on the Et-GaAs-30min film, coming from the high density of surface native oxides and totally impeding the charge transfer kinetics at the interface.This augmentation in the area of the Ti 3+ peak signifies that either a large amount of Ti2O3(+3) is formed, or that some mixed oxide structure of Ti-O-Ni is formed after co-catalyst deposition. 4,5ese defect states in the crystalline TiO2 are mainly responsible for the hole transportation toward the surface to drive the significant PEC water oxidation.To understand the quantitative generation of oxygen vacancies (Ov), electron paramagnetic resonance (EPR) or electron spin resonance (ESR) spectroscopy is recorded to survey the presence of oxygen vacancies or Ti 3+ , with paramagnetic species containing unpaired electrons.Here, Et-GaAs/TiO2/Ni-Pi photoanode films were prepared under different annealing temperature conditions, and their EPR characters were compared at room temperature, as shown in Supplementary Fig.As shown in Supplementary Fig. 13, the FE(%) of GaAs film decreased over time, subsequently corresponding to the photo-corrosion effect.In the case of photo-corrosion, the excited electrons or holes can react with the electrolyte or the semiconductor itself, leading to the formation of reactive byproducts such as hydroxyl radicals or other oxygen species or defects that can reduce the PEC performance.These side reactions (e.g., non-conductive native oxide or nonfaradic charging process at the interface etc.) induce to a side current that competes with the desired photocurrent, leading to lower FE.Also, this dark current can be caused by the formation of a potential barrier at the interface between the corroded and non-corroded parts of the material.The potential barrier acts as a barrier for the electron flow, making the photocurrent pathway around the corroded area.
In the case of Et-GaAs film, the FE is significant at the initial reaction time due to the high surface area effect.As time is going, the etched GaAs photoanodes are vulnerable to photocorrosion, which can degrade their PEC performance over time.Etching can increase the surface area of the GaAs photoanode, which provides more sites for the electrolyte to react with the photoelectrode, leading to an increased susceptibility to photo-corrosion.Therefore, minimizing the photo-corrosion effect is important for improving the PEC performance and stability.However, there is no meaning degradation observed in the photoelectrodes such as Et-GaAs/TiO2, and Et-GaAs/TiO2/NiPi photoanodes due to the surface passivation effect, surely demonstrating the stable PEC performance and co-catalyst boosts the rapid charge transfer reaction at the surface interface.
formation of Ga-P after the long-term stability test.To delicately survey the P element after longterm stability testing, FIB-TEM analysis integrated with EDX mapping and in-depth line mapping was performed, as shown in Figure 5(f).We found that the PO4 )), the pure GaAs and Et-GaAs are well fitted with the resolved peaks.The peaks at binding energies of 530.87 eV and 531.92 eV are attributed to native oxides, such as Ga-O and As-O, and non-lattice oxygen in bare GaAs film.The peaks at binding energies of 529.91 eV, 531.03 eV, and 532.22 eV are attributed to lattice oxygen in As2O3, Ga-O, or As-O and nonlattice oxygen, ascribed to metal bounded hydroxides in Et-GaAs film.This confirms the formation of a native oxide layer in the etched GaAs film during the electrochemical etching process.

Supplementary Figure 5 .
photocurrent decay is observed in the first 5 h of the stability testing at 0.6 and 1.23 VRHE, and the pronounced decay is 61% and 51%, respectively.Meanwhile, the observed decay at 0 VRHE is about 72%, proving that the bare GaAs corrosion rate depends highly on the applied potential and that the GaAs is a sensitive material under PEC working conditions.Similarly, the dissolution rate as a function of the prolonged time (25 h and 50 h) was summarized in Supplementary Fig.5(b), revealing that the dissolution of GaAs in low potential is slower, middle potential is moderate, and high potential is extreme.However, after 50 h of testing, all amperometric (i-t) curves reached the same value, and GaAs completely detached from top to bottom.To understand the dissolution mechanism in depth, the stability test was carried out at different light intensities at a constant potential of 1.23VRHE.The inset of Supplementary Fig.5(b)shows the GaAs dissolution rate under 1-, 0.75-, and 0.5-sun illumination, respectively.A lower light intensity (0.5-sun) led to better stability up to ~25 h, whereas a faster dissolution rate was observed under 1-sun and 0.75-sun intensity.The lower intensity leads to limited PEC performance due to the deficient energy carrier separation and increased kinetic overpotential for the water oxidation.Also, the surface morphology and structure of the bare GaAs film after the long-term stability testing at different applied potentials of 0, 0.6, and 1.23 VRHE were also probed and are displayed in Supplementary Fig.5(c-f).A strong alkaline electrolyte with a high potential to induce some cracked or rougher

7 . 7 .
Then, all films, including GaAs, Et-GaAs/TiO2 (250°C)/Ni-Pi, Et-GaAs/TiO2 (300°C)/Ni-Pi, and Et-GaAs/TiO2(350°C)/Ni-Pi, showed strong resonance signals.Considering that the intensity variation of the EPR signal represents the presence of different magnitudes of oxygen vacancies, the EPR signal of Et-GaAs/TiO2 (300°C)/Ni-Pi film exhibits a more intense and broader peak than the others.The most intense peak, with g factor at 2.002 and 1.983, implies that Ov and electrons trapped on oxygen vacancies were closely associated with Ti 3+ on oxygen vacancies.The results demonstrate that Ov at the Et-GaAs/TiO2-300°C film possesses the optimum concentration, contributing to the superior PEC performance (Figure4(a)).Furthermore, the core-level Ni2p spectrum of the Et-GaAs/TiO2/Ni-Pi sample shows the BE peaks at 856.23, 857.40, 873.82, and 875.74 eV, consistent with those of Ni2p3/2 and Ni2p1/2 in Ni3(PO4)2, assigned to the two spin-orbit doublets characteristic of Ni 2+ and Ni 3+ and two shakeup satellite peaks (labeled as Sat. 1 and Sat. 2).The peaks can interact with phosphate and hydroxide ions.In addition, the higher binding energy of 857.40 and 875.74 eV in the Ni2p spectra can be assigned to the core-level peaks of Ni 2+ cations, indicating the presence of Ni(OH)2 in the hydrated Ni3(PO4)2 structure.6,7The core-level P2p spectra reveals the binding energy peaks at 134.48 and 135.58 eV, which were well fitted to 2p1/2 and 2p3/2 doublets, as displayed in Supplementary Fig.6(e), ascribed to interaction of the P-O bonding in Supplementary Figure Electron paramagnetic resonance (EPR) of GaAs, Et-GaAs/TiO2 (250°C)/Ni-Pi, Et-GaAs/TiO2 (300°C)/Ni-Pi, and Et-GaAs/TiO2 (350°C)/Ni-Pi films.Supplementary Figure 13.O2 evolution rate corresponding the FE (%) of GaAs, Et-GaAs, Et-GaAs/TiO2 and Et-GaAs/TiO2/NiPi photoanode as a function of time.

Table 2 .
3-ions combined with Ga to form a Ga-O-P layer on the surface after long-term stability testing.Furthermore, in the case of the O1s Summary on the GaAs based PEC performance in terms of fabrication method, photocurrent density (Jph), electrolyte, and stability.