Enhanced Photocatalytic Hydrogen Evolution by Loading Cd0.5Zn0.5S QDs onto Ni2P Porous Nanosheets

Ni2P has been decorated on CdS nanowires or nanorods for efficient photocatalytic H2 production, whereas the specific surface area remains limited because of the large size. Here, the composites of Cd0.5Zn0.5S quantum dots (QDs) on thin Ni2P porous nanosheets with high specific surface area were constructed for noble metal-free photocatalytic H2 generation. The porous Ni2P nanosheets, which were formed by the interconnection of 15–30 nm-sized Ni2P nanoparticles, allowed the uniform loading of 7 nm-sized Cd0.5Zn0.5S QDs and the loading density being controllable. By tuning the content of Ni2P, H2 generation rates of 43.3 μM h− 1 (1 mg photocatalyst) and 700 μM h− 1 (100 mg photocatalyst) and a solar to hydrogen efficiency of 1.5% were achieved for the Ni2P-Cd0.5Zn0.5S composites. The effect of Ni2P content on the light absorption, photoluminescence, and electrochemical property of the composite was systematically studied. Together with the band structure calculation based on density functional theory, the promotion of Ni2P in charge transfer and HER activity together with the shading effect on light absorption were revealed. Such a strategy can be applied to other photocatalysts toward efficient solar hydrogen generation.


Background
As an efficient strategy to produce H 2 by utilizing solar energy, photocatalytic hydrogen production has attracted extensive attention since TiO 2 was reported as a photocatalyst in 1972 [1]. Compared with TiO 2 , Cd x Zn 1−x S shows excellent visible-light driven catalytic activity because of the narrower band gap and good photochemical stability. A H 2 production rate as high as 1097 μM h − 1 g − 1 has been achieved by using Cd 0.5 Zn 0.5 S as photocatalyst [2], which composition has been proven to be the optimum for photocatalytic property. To decrease carrier recombination and prompt carrier separation for hydrogen evolution reaction (HER), noble metals such as Pt, Co-Pt, Ru, Au, and Pd have been used as cocatalysts [3][4][5][6][7][8]. For example, when co-catalyzed with Co-Pt, the photocatalytic H 2 generation rate of Cd 0.5 Zn 0.5 S quantum dots (QDs) could be increased by 4.7-folds [4]. A H 2 production as high as~6.3 mM h − 1 mg − 1 was achieved when CdZnS was combined with Au [9]. However, the high cost of noble-metals greatly limits the future application in large scale, which makes the non-precious co-catalysts to be good candidates of precious ones for photocatalytic H 2 generation.
Here, a reverse structure of Cd 0.5 Zn 0.5 S QDs on Ni 2 P nanosheet arrays was synthesized by thermal solution method for enhanced photocatalytic H 2 generation. A hydrogen generation rate of 700 μM h − 1 (with 100 mg feeding catalyst) and a solar to hydrogen efficiency (STH) of 1.5% were achieved at 1.5 wt% of Ni 2 P. The effect of Ni 2 P on the H 2 generation rate, optical, and electrochemical property of the composite was systematically studied. Moreover, the band structure of Ni 2 P was calculated based on density functional theory, together with the photo-electrochemical property, the detailed role of Ni 2 P for the H 2 generation was revealed.

Synthesis of Co-catalyst
Firstly, 20 mL deionized water containing 2.61 g nickel nitrate and 2.52 g hexamethylenetetramine was transferred to a Teflon autoclave and heated at 120°C for 10 h for the formation of NiOOH. After cooled down to room temperature, the NiOOH product was washed by alcohol and deionized water via centrifugation at 2000 rpm for three times and each time for 5 min. Then, a mixture of 0.22 g NiOOH and 0.44 g sodium hypophosphite was put into a tube furnace and heated at 500°C for 2 h for phosphorizing. When it naturally cooled down to room temperature, black Ni 2 P powder was obtained and collected.
Synthesis of Ni 2 P-Cd 0.5 Zn 0.5 S Nanocomposites To prepare Ni 2 P-Cd 0.5 Zn 0.5 S composite, 100 mg Ni 2 P powder was dispersed into 20 mL ethanol via ultrasonic processing for 1 h. Then x mL (x = 0.48, 0.96, 1.4, 3, 5) well-dispersed Ni 2 P solution was added into a 20 mL ethylene glycol solution containing 272.6 mg ZnCl 2 and 456.7 mg CdCl 2 •2.5H 2 O, and was heated to 170°C with continuous stirring under nitrogen protection. After the addition of 20 mL ethylene glycol solution dissolving 960.7 mg Na 2 S•9H 2 O, the solution was heated to 180°C and held for 1 h for the growth of Cd 0.5 Zn 0.5 S on Ni 2 P.
Finally, the samples were washed by alcohol and deionized water respectively for three times. By weighing the final xNi 2 P-Cd 0.5 Zn 0.5 S composites, the weight percents (wt%) were determined to be 0.5 (x = 0.48), 1 (x = 0.96), 1.5 (x = 1.4), 3 (x = 3), 5 (x = 5). As a comparison, pure Cd 0.5 Zn 0.5 S QDs were synthesized via the similar method except the addition of Ni 2 P.

Morphology, Structure, and Optical Properties Characterization
The morphology, microstructure, and composition were characterized by field emission scanning electron microscopy (FESEM, JSM-7100F, JEOL) and transmission electron microscopy (TEM, FEI Tecnai 20) equipped with scanning transmission electron microscopy (STEM) and energy dispersive X-ray spectroscopy (EDX). Powder X-ray diffraction (XRD) patterns were recorded on a Bruker AXS D8 X-ray diffractometer with Cu Kα (λ = 1.54056 Å). Elemental composition, chemical, and valence states were studied by (valence band) X-ray photoelectron spectroscopy (XPS) measurements (XPS, Escalab 250Xi) with Al Kα radiation. UV-Vis absorption was investigated by an UV-Vis spectrophotometer (UV-3600, Shimadzu) equipped with an integrating sphere device, and the weight/volume ratio of sample to deionized water was kept at 1 mg/10 mL. Photoluminescence (PL) measurements were carried out on a 7000 FL spectrophotometer (Hitachi, F7000) with an excitation wavelength of 400 nm. Before the PL measurements, pure Cd 0.5 Zn 0.5 S QDs and the composites were well dispersed in ethanol, and the concentration of Cd 0.5 Zn 0.5 S was maintained at 0.5 mg/mL for all the samples.

Linear Sweep Voltammetry (LSV) and Electrochemical Impedance Spectra (EIS) Measurements
LSV measurements were conducted in 1 M NaOH electrolyte (pH = 14) in an electrochemical work station (CHI 760E, CH Instruments) with a typical threeelectrode configuration. A Pt foil and a saturated Ag/ AgCl were used as the counter and reference electrode, respectively. The potentials were converted to those vs reversible hydrogen electrode (RHE) by the equation [39]. Electrochemical impedance spectra (EIS) measurements were carried out in darkness at 0.5 V vs RHE with an amplitude of 5 mV and the electrolyte of 0.35 M Na 2 SO 3 and 0.25 M Na 2 S aqueous solution by using a similar three-electrode system. The working electrode was made via spreading2 mg product (dispersed in 5 mL ethanol) over 4 cm 2 area FTO substrate and dried at 70°C for 5 h. The frequency range was kept within 0.1 Hz~100 kHz, and the spectra were analyzed by the Z-View program (Scribner Associates Inc.).

Photocatalytic (PC) H 2 Generation
Before H 2 production, the photocatalysts with different mass (1, 5, and 10 mg) were dispersed in a sealed quartz reactor (volume 40 mL, 5 cm × 5 cm × 1.6 cm) with 15 mL 0.75 M Na 2 S and 1.05 M Na 2 SO 3 aqueous solution. After degassing for 30 min by nitrogen, the photocatalytic experiment was performed under the irradiation of a 300 W Xe (PLS-SXE300/300UV, Perfect Light) lamp with a cut-off filter of 420 nm and an incident power of 300 mW/cm 2 . The catalytic solution was kept continually stirring during the whole PC experiment. In every hour, 1-mL gas production was collected and analyzed by a gas chromatograph (GC-2018, Shimadzu, Japan, TCD). Further cycling stability experiment was performed under the same condition. Paralleling experiments with the feeding dosage of photocalysts from 15 to 100 mg were conducted in 100 mL electrolyte of Na 2 S and Na 2 SO 3 in a larger reactor (volume 150 mL) under the same illumination. The solar to hydrogen efficiency (STH) was calculated by the flowing equation:

Computational Methods
The energy and electronic properties of bulk Ni 2 P were calculated using density functional theory (DFT) method. Vienna Ab-initio Simulation Package (VASP) [40] was adopted during the calculations with the projector augmented wave pseudo potentials (PAW) [41], and the Perdew-Burke-Ernzerhof type (PBE) generalized gradient approximation (GGA) [42] exchange-correlational functional methods. A Brillouin zone with a 9 × 9 × 9 Monkhorst−Pack Γpoint grid [43], a kinetic energy cut off with 450 eV, and an energy criterion of 10 − 6 eV were Fig. 1 Morphology, crystalline property, and chemical states of Ni 2 P-Cd 0.5 Zn 0.5 S composites (1.5 wt% Ni 2 P). a-b Low and high (inset) magnification SEM images of Ni 2 P before and after the loading of Cd 0.5 Zn 0.5 S, c XRD pattern of Ni 2 P and Ni 2 P-Cd 0.5 Zn 0.5 S, d-f XPS fine and survey spectra of Ni 2 P-Cd 0.5 Zn 0.5 S composite applied for geometric optimization until the residual forces were converged to less than 0.01 eV/Å. The bulk model of hexagonal Ni 2 P with P-62M symmetry was taken into account. After fully structure optimized, the lattice parameter of Ni 2 P (a = b = 5.86918 Å, and c = 3.37027 Å) can be obtained, which is well consistent with the reported values [44]. Figure 1a, b show the morphology of Ni 2 P before and after the composition with Cd 0.5 Zn 0.5 S QDs (Ni 2 P wt%: 1.5%). Pure Ni 2 P has a flower-like morphology which is composed of many crossed nanosheets with the thickness less than 20 nm and planar size from several tens nanometer to micrometer scope. From the XRD pattern of pure Ni 2 P in Fig. 1c, diffraction peaks of (111), (201), (210), and (300) planes can be clearly observed at 40.7°, 44.6°, 47.4°, and 54.2°, respectively, which correspond to hexagonal Ni 2 P (JCPDF no. 89-2742). After loaded by Cd 0.5 Zn 0.5 S QDs, the surface of the nanosheets become rather rough, and plenty of nanoparticles with size less than 10 nm can be distinguished on the pristine Ni 2 P skeleton. At the same time, the XRD refraction peaks of Cd 0.5 Zn 0.5 S (JCPDF no. 89-2943) (100), (002), (101), and (110) planes can be clearly found at 26.0°, 27.8°, 29.6°, and 45.9°, respectively [6,45], while the diffraction signal of Ni 2 P is greatly depressed because of the low weight ratio (1.5 wt%) of Ni 2 P to Cd 0.5 Zn 0.5 S. The coexistence of Cd 0.5 Zn 0.5 S and Ni 2 P was demonstrated by the X-ray photoelectron spectrometer (XPS) fine and survey spectra in Fig. 1d-f. Except the oxygen and carbon signals arising from the air absorption, only Ni, P, Cd, Zn, and S can be detected, which rules out the possibility of other impurities. The peaks at 855.5 and 873.9 eV can be assigned to Ni 2p 3/2 and 2p 1/2 , respectively, and the peak of P 2p 3/2 can be found at 133.6 eV [16,46]. Concurrently, the doublet peaks of Zn 2p, Cd 3d, and S 2p suggest the bivalent Zn 2+ , Cd 2+ , and S 2− from Cd 0.5 Zn 0.5 S QDs [3,34,47]. In brief, the growth of Cd 0.5 Zn 0.5 S on Ni 2 P nanosheets has been established for the formation of Ni 2 P-Cd 0.5 Zn 0.5 S nanocomposites.

Results and Discussion
The microstructure and elemental composition of the samples were further investigated by TEM-related techniques. From the different magnification TEM images of pure Ni 2 P (Fig. 2a, b), the nanosheets are porous and composed of cross-linked irregular nanoparticles with size of~15-30 nm. The selected area electron diffraction pattern (SAED) in Fig. 2c shows the diffraction ring of Ni 2 P (111), (201), (210), and (300) planes. The diffractive signals of high-index planes such as (222), (402), Fig. 2 Microstructure of Ni 2 P and Ni 2 P-Cd 0.5 Zn 0.5 S composite. a-c and d-f Different-magnification TEM images and SAED pattern of Ni 2 P and Ni 2 P-Cd 0.5 Zn 0.5 S, the inset f is EDX spectrum, where the yellow and white dash lines denote Cd 0.5 Zn 0.5 S and Ni 2 P, respectively. g High-angle annular dark field (HAADF)-STEM image, and h-l the corresponding EDX mappings of Ni 2 P-Cd 0.5 Zn 0.5 S composite and (420) can also be detected due to the strong multiscattering of the high-energy electrons. After composited with Cd 0.5 Zn 0.5 S, the intercrossed Ni 2 P nanosheets were covered by plenty of smaller nanoparticles with size of7 nm (Fig. 2d). The EDX spectra (inset, Fig. 2f) clearly shows the signal of Ni, P, Cd, Zn, and S, indicative of the coexistence of Ni 2 P and Cd 0.5 Zn 0.5 S. From the SAED pattern (Fig. 2f), strong diffractive rings of Cd 0.5 Zn 0.5 S (002), (110), and (200) planes (denoted by yellow dash lines) can be clearly distinguished along with the weak signals of Ni 2 P (300), (402), and (420) (marked by white dash lines), suggesting the good composition of Ni 2 P with QDs. It is noticeable that Ni 2 P (300) ring overlaps with Cd 0.5 Zn 0.5 S (110) and (200) planes, making it hard to be distinguished. The high-resolution TEM image of Ni 2 P-Cd 0.5 Zn 0.5 S sample in Fig. 2e further shows the lattice fringes with spacing of 0.34 and 0.22 nm, which corresponds to the Cd 0.5 Zn 0.5 S (002) and Ni 2 P (111) crystal planes, respectively. The elemental EDX mappings (Fig. 2h-l) taken from the region shown by the high-angle annular dark field (HAADF) image (Fig. 2g) exhibit that Ni, P, Cd, Zn, and S are distributed uniformly among the sample, further demonstrating the successful composition of Cd 0.5 Zn 0.5 S QDs with the porous Ni 2 P nanosheets. Figure 3a shows the H 2 evolution rate of Ni 2 P-Cd 0.5 Zn 0.5 S nanocomposites varied with the content of Ni 2 P at the feeding dosage of 1 mg in a 40 mL reactor. Pure Cd 0.5 Zn 0.5 S shows a photocatalytic H 2 evolution rate of 12.6 μM h − 1 mg − 1 , and pure Ni 2 P shows negligible hydrogen generation. With the addition of Ni 2 P, the photocatalytic activity of the Ni 2 P-Cd 0.5 Zn 0.5 S composites has been obviously enhanced and reaches the highest value of 43.3 μM h − 1 mg − 1 at 1.5 wt% Ni 2 P, nearly 3.4 times higher than pure Cd 0.5 Zn 0.5 S. Further addition of Ni 2 P (≥ 3 wt%) will result in fast degradation of property, and the H 2 evolution rate is less than pure Cd 0.5 Zn 0.5 S when Ni 2 P increases to 5 wt%. Such a nonlinear behavior suggests that there exist an optimum Ni 2 P content, namely, an appropriate loading density of Cd 0.5 Zn 0.5 S on Ni 2 P for the photocatalytic property. At the same time, the stability of 1.5 wt% Ni 2 P-Cd 0.5 Zn 0.5 S was studied by cycling test (Fig. 3b). During four successive cycles that lasted for totally 16 h, the H 2 generation maintained relative stable with negligible degradation, indicating the good photocatalytic stability of the composite.
The effect of the amount of catalyst on STH efficiency and H 2 generation was systematically studied (Fig. 3c-d)   Fig. 3 Photocatalytic property of Ni 2 P-Cd 0.5 Zn 0.5 S composites. a Photocatalytic hydrogen generation at different wt% of Ni 2 P and b the cycling test of the composite with 1.5 wt% of Ni 2 P tested in a small reactor (40 mL, 1.0 mg photocatalyst). c Hydrogen production rate and solar to hydrogen efficiency (STH) at various amount of photocatalyst. The tests for the photocatalyst of dosage from 15 to 100 mg were carried out in a 150 mL reactor, and those of dosage from 1 to 10 mg were in a 40 mL reactor. d The hydrogen generation rate for 1 and 100 mg composite samples (1.5 wt% Ni 2 P) for 1.5 wt% Ni 2 P-Cd 0.5 Zn 0.5 S sample. Two typical reactors with volume of 40 and 150 mL were adopted at the same illumination power density. When tested in the smaller reactor (40 mL), though both the STH and H 2 generation rate increase with the catalyst's dosage from 1 to 10 mg, the increased step is far less than that of the dosage. The STH and H 2 generation rate are only 0.45% and 166 μM h − 1 when the dosage of the catalyst increased to 10 mg, nearly 3.8 times of the 1 mg sample. For the larger reactor (150 mL), distinct increase in STH and H 2 generation can be found with the dosage increased from 15 to 100 mg, and a 1.53% STH and a 700 μM h − 1 of H 2 generation can be achieved at the dosage of 100 mg, nearly 3.1 times of the 15 mg catalyst. Considering that the incident light has longer path when it passes through a deeper reactor, such a result shows that larger reactor will be more beneficial for the utilization of the incident light. However, the STH efficiency will be saturated once the dosage increased to about 100 mg, suggesting there exists an optimum dosage for the light utilization. The optimum H 2 generation rate is superior than CdZnS QDs-2D g-C 3 N 4 microribbons (H 2 generation rate 33.4 mM h − 1 g − 1 ) [10], Cd 0.1 Zn 0.9 S nanoparticles-carbon nanotubes (rate: 1563 μM h − 1 g − 1 ) [11], a sandwich-structured C 3 N 4 / Au/CdZnS photocatalyst (rate 6.15 mM h − 1 g − 1 ) [9], and CdS QDs-sensitized Zn 1−x Cd x S solid solutions (rate 2128 μM h − 1 g − 1 ) [48].
To reveal the mechanism for the enhanced photocatalytic property and detailed role of Ni 2 P, both the optical and electrochemical property of pure Ni 2 P, Cd 0.5 Zn 0.5 S, and the composites were studied by Fig. 4. From the absorption spectra (Fig. 4a), pure Cd 0.5 Zn 0.5 S exhibits an absorption edge at 506 nm, corresponding to the band gap of 2.45 eV [13,49]. For pure Ni 2 P (the inset), wide absorption over the whole visible range can be found. After the composition, besides the absorption in range < 506 nm, obvious tails over the visible wavelength > 506 nm can be found, which can be attributed to the contribution from Ni 2 P. As the visible absorption in longer wavelength increases with Ni 2 P, the composite shows reduced absorption of Cd 0.5 Zn 0.5 S (< 506 nm). At the same time, the photoluminescence spectra (Fig. 4b) exhibit that pure Cd 0.5 Zn 0.5 S has intensive band edge luminescence at~620 nm when excited at the wavelength of 400 nm. After composition, it will be degraded gradually with the addition of Ni 2 P. Considering that higher content of Ni 2 P will induce more Ni 2 P/Cd 0.5 Zn 0.5 S interfaces which help to enhance charge transfer and suppress charge recombination, the decrease of PL intensity can be understood by the reduced carrier recombination and enhanced charge transfer at the Ni 2 P/ Cd 0.5 Zn 0.5 S interface.
The effective role of Ni 2 P in prompting charge transfer can also be reflected by the EIS spectra depending on Ni 2 P content (Fig. 4c). As shown by the equivalent Fig. 4 The effect of Ni 2 P content on the optical and electrochemical properties of Ni 2 P-Cd 0.5 Zn 0.5 S composite. a UV-Vis absorption spectra (inset pure Ni 2 P), b photoluminescence spectra, and c EIS spectra. d LSV curve and EIS (inset) spectrum of pure Ni 2 P circuit (inset, Fig. 4c), the charge transfer resistance (Rct) at catalyst/electrolyte interface can be evaluated by the semicircle radius of the Nyquist plots based on R-C equivalent circuit. The equivalent series resistance (ESR) can be obtained from the intersection of the curve and the real resistance (Z') axis, while the charge-transfer resistance (Rct) corresponds to the width of the semicircle plotted at higher frequencies. The R CT of pure Cd 0.5 Zn 0.5 S is 17,320 Ω, indicative of its semiconductor nature. After the composition with 1, 1.5, and 3 wt% Ni 2 P, R CT decreases gradually to 8432, 7721, and 5473 Ω, respectively, suggesting the enhancement of Ni 2 P in the electrical conductivity. Indeed, Ni 2 P has been considered as a good electrocatalyst toward HER [44,50,51]. From the LSV curve of pure Ni 2 P on Ni foam shown in Fig. 4d, the Ni 2 P has good HER activity with overpotentials of 84 mV and 201 mV to attach the current density of 10 and 50 mA/cm 2 (without iR-correction), respectively. The EIS spectrum (inset Fig. 4d) exhibits that Ni 2 P has a very low R CT (~7.3 Ω), indicating the metallic character of Ni 2 P. Therefore, Ni 2 P can not only increase the electrical conductivity at Cd 0.5 Zn 0.5 S/Ni 2 P interface, but also supply effective active sites for HER, then leading to enhanced photocatalytic property of the composite.
Considering that the addition of Ni 2 P decreased the absorption at wavelength < 506 nm, it is necessary to demonstrate whether the light absorption of Ni 2 P can be utilized to generate hydrogen. The band structure of Ni 2 P was then studied by DFT calculation. Figure 5a, b presents the ball and stick model of bulk Ni 2 P and the calculated band structure. From Fig. 5b, no band gap can be detected, suggesting the metallic characteristic of Ni 2 P, which agrees well with the above EIS result. This indicates that the photoelectrons are mainly attributed to the photo-excitation of Cd 0.5 Zn 0.5 S rather than Ni 2 P. Moreover, the Fermi level of Ni 2 P (obtained from out car file) locates at 1.03 V vs. NHE, much lower than the conductive band minimum (CBM) level (− 1.04 V vs NHE) of Cd 0.5 Zn 0.5 S QDs [13].
Accordingly, the schematic mechanism was demonstrated for the photocatalytic H 2 evolution of the composite by Fig. 5c. The location of Fermi level of Ni 2 P makes it energetically favorable for the transfer of photo-generated electrons from Cd 0.5 Zn 0.5 S to Ni 2 P, then prompts the separation of photo-excited electrons and holes at the interface, resulting in the depression of charge recombination. Concurrently, H 2 will evolve efficiently at the active sites of Ni 2 P due to the good HER Fig. 5 The band diagram and charge separation and transfer mechanism for the photocatalytic H 2 evolution. a Top views of the ball and stick model of (001) surface-terminated bulk Ni 2 P. b Calculated band structure of Ni 2 P where the red dash line represents Fermi level. c Schematic mechanism illustrating the charge separation and transfer for the photocatalytic H 2 generation activity and large specific surface area of the composites. The positive roles of Ni 2 P in charge transfer and HER activity will dominate at the lower content of Ni 2 P (≤ 1.5 wt%). When the content surpasses 1.5 wt%, the shading effect of Ni 2 P in light absorption will overcome the positive aspect, leading to the degradation of H 2 generation. An optimum photocatalytic property will be achieved at 1.5 wt% Ni 2 P when the two effects reach a balance.

Conclusions
A reverse structure of Cd 0.5 Zn 0.5 S QDs on Ni 2 P porous nanosheets were fabricated for efficient photocatalytic H 2 production. The Ni 2 P porous nanosheets were composed of 15-30-nm-sized nanoparticles that allows the effective loading of 7-nm-sized Cd 0.5 Zn 0.5 S QDs. As the charge separation and transfer property is enhanced with the addition of Ni 2 P from 0 to 5 wt%, a competitive shading effect that unbeneficial for the light absorption of Cd 0.5 Zn 0.5 S is induced. An optimum photocatalytic H 2 generation of 43.3 μM h − 1 (dosage 1 mg) will be achieved at 1.5 wt% Ni 2 P. Based on the optimum content, the photocatalytic dependence on feeding dosage of catalyst shows that the STH efficiency will reach the highest value of 1.5% at the dosage of 100 mg. The high HER activity and band structure of Ni 2 P were revealed, confirming the effective role of Ni 2 P in prompting photocatalytic H 2 evolution dynamics from both experimental and theoretical aspects. The heterostructure of Cn 0.5 Zn 0.5 S QDs-Ni 2 P porous nanosheets can not only help to prompt the photo-excited charge separation and transfer, but also speed up the dynamics of hydrogen evolution reaction via the co-catalytic role of N i2 P, thus enhances the photocatalytic hydrogen generation property. Such a method can be applied to other catalysts toward efficient photocatalytic property.