Controlled Synthesis and Photoelectrochemical Performance Enhancement of Cu_2−xSe Decorated Porous Au/Bi₂Se₃ Z-Scheme Plasmonic Photoelectrocatalyst

Abstract

In this paper, uniform Cu_2−xSe-modified Au/Bi₂Se₃ hybrid nanoparticles with porous shells have been prepared through a cation exchange method. Bi₂Se₃/Cu_2−xSe Z-scheme heterojunction is introduced onto Au nanocube by replacing Bi³⁺ with Cu²⁺. Owing to the effective coupling between Au core and semiconductor shells, Au/Bi₂Se₃/Cu_2−xSe hybrids present a broad and strong plasmon resonance absorption in the visible band. More intriguingly, the carrier lifetime of Au/Bi₂Se₃/Cu_2−xSe hybrid photoelectrodes can be further tailored with corresponding Cu_2−xSe content. Through parameter optimization, 0.1-Au/Bi₂Se₃/Cu_2−xSe electrode exhibits the longest electron lifetime (86.03 ms) among all the parallel samples, and corresponding photoelectrochemical performance enhancement is also observed in the tests. Compared with that of pure Bi₂Se₃ (0.016% at 0.90 V vs. RHE) and Au/Bi₂Se₃ (0.02% at 0.90 V vs. RHE) nanoparticles, the maximum photoconversion efficiency of porous Au/Bi₂Se₃/Cu_2−xSe hybrid photoanodes increased by 5.87 and 4.50 times under simulated sunlight illumination, attributing to the cooperation of Z-scheme heterojunction and plasmon resonance enhancement effects. All the results indicate that Au/Bi₂Se₃/Cu_2−xSe porous hybrids combine eco-friendliness with excellent sunlight harvesting capability and effectively inhibiting the charge recombination, which provide a new idea for efficient solar-driven water splitting.

1. Introduction

With the consumption of non-renewable fossil fuel and the resulting environmental pollution, energy and environmental issues are globally important topics [1,2]. Over several decades, enormous efforts have been made to explore new materials and technologies for converting solar energy into other energies efficiently, owing to the solar energy’s clean, abundant and irreplaceable property [3,4,5,6,7]. Photoelectrochemical (PEC) water splitting based on semiconductor photoactive materials is well recognized as one of the ideal schemes since the initial report on PEC water splitting in 1972 [8,9,10,11]. Metal chalcogenides nanocrystals have attracted tremendous attention due to their tunable and suitable conduction band edges for facile hydrogen evolution reaction [12,13,14,15]. Among them, Bi₂Se₃ is one of the promising n-type semiconducting material belonging to VB-VIA group, with an appropriate band gap (0.3–2.3 eV), unique topological surface states and eco-friendliness. It has been receiving considerable attention for various applications including photocatalysts, photoelectrocatalysts, thermoelectric and photoelectric devices [16,17,18,19].

Unfortunately, the inherent shortcomings of semiconductors, such as low photoelectric conversion efficiency, high charge transfer resistance, fast recombination speed, hinder their applications in PEC water splitting. In order to overcome those bottlenecks, various strategies have been proposed, such as doping, heterojunction and plasmonic enhancement [20,21,22,23,24]. For instance, N-substituted bismuth selenide material demonstrates a superior performance for degradation of multi-pollutants and the ciprofloxacin and phenol removal rates reach 88.8 and 83.6%, respectively, under irradiation of simulated solar for 2 h [25]. Nonstoichiometric Cu_2−xSe, as a p-type and self-doped semiconductor, possesses not only excellent optical properties as its bandgap lies in the range of 1.4–2.2 eV, but also ultrahigh electrical conductivity tailored by x value. These interesting properties creates its importance towards its application in natural sunlight-driven photocatalysis and photoelectrocatalysis [26,27]. In addition, based on two semiconductors with unequal band gaps by physical contact each other, construction of heterogeneous photocatalyst systems is also a set of effective schemes to achieve highly efficient solar energy conversion to chemical energy [28,29,30,31,32]. The narrow band gap semiconductor is introduced into the host semiconductor, which can broaden the practical application of solar spectrum [33,34,35]. Through proper choice of both the semiconductor with respect to energy level alignment and CB/VB edge position, a type-II or Z-scheme heterostructure can be further constructed to improve the transport performance of photogenerated carrier [22,29,36,37]. For example, Bi₂Se₃ nanoflowers were incorporated onto the TiO₂ electrode in Subramanyam’s work, the formation of type-II heterojunction is beneficial to improve the PEC performance of hybrid electrode [38]. In Qu’s report, Cu_2−xSe/Bi₂Se₃ hybrids exhibited the enhanced near-infrared light harvest and photoredox abilities through constructing Z-scheme heterostructure [36]. Beside those, plasmon nanoparticles are often designed as light-trapping units and reaction hot spots in hetero photocatalysts for highly efficient harvesting and utilizing of solar energy. Great efforts have been made to explore plasmon-mediated solar energy conversion mechanisms based on metal-semiconductor nanostructures, such as radiative relaxation of surface plasmon resonance (SPR), hot electron transfer and plasmon-induced resonant energy transfer [21,39,40,41,42,43,44]. Recently, under the resonant surface plasmon excitation of Au nanoparticles, the decently enhanced photoelectrocatalytic performances of Au/Bi₂Se₃ nanoflowers for the hydrogen evolution reaction was reported by Li and co-workers [43]. In addition, two kinds of plasmonic Au and Cu_2−xSe nanocrystals were ingeniously assembled into one system in our previous report, achieving dual-plasmon resonance coupling and enhancement of their photocatalytic performance [44].

However, there are only few studies on integrating the two plasmon nanostructures into Bi₂Se₃ nanosystem and their cooperation mechanism of enhanced PEC performance, owing to the limitations of synthetic technology. Those provide powerful motivation for our research in this paper. Herein, uniform Cu_2−xSe decorated Au/Bi₂Se₃ hybrid nanoparticles with porous shells are prepared by using a cation exchange method. As an eco-friendly photoelectrocatalysts, the Cu_2−xSe-dependent photoelectrochemical performances of Au/Bi₂Se₃/Cu_2−xSe photoelectrodes have been further investigated by PEC tests. Based on the synergistic effect of plasmonic nanostructures and semiconductor heterojunctions, the designed Z-scheme plasmonic photoelectrocatalyst exhibited excellent photoelectrochemical performances.

2. Results and Discussion

In this work, employed Au nanocubes (NCs) as the seeds, the porous Au/Bi₂Se₃/Cu_2−xSe nanoparticles were fabricated through a simple three-step method, as illustrated in Figure 1a. Step one is the synthesis of the Au/Se Janus-like nanostructures. Using hexadecyltrimethylammonium bromide (CTAB) as the stabilizer, the Au seeds are coated with an amorphous Se layer by reducing SeO₂ with the ascorbic acid (AA), forming Janus-shape Au/Se nanostructures. Step two is the preparation of porous Au/Bi₂Se₃ hybrid nanostructures with the help of glycine transporter. The detailed mechanism has been discussed in our previous work [21]. Last step is the formation of Bi₂Se₃/Cu_2−xSe heterojunction. Cu²⁺ can be reduced to Cu⁺ by reducing agent AA, and then cation exchange reaction with Bi³⁺ ions in Au-Bi₂Se₃ solution can be performed to produce Cu_2−xSe wetting layer on Au/Bi₂Se₃. Additionally, the epitaxial growth of Cu_2−xSe shells inevitably occurs due to the presence of excess Se. The amount of Cu_2−xSe can be further regulated by the initial Cu²⁺ precursor concentration (in Figure S1). Here, Cu²⁺ precursor concentration is set as x (x = 0.01 M, 0.1 M, and 0.4 M), and the obtained final product is labeled as x-Au/Bi₂Se₃/Cu_2−xSe. Taking 0.1-Au/Bi₂Se₃/Cu_2−xSe as an example, the corresponding synthesis process are characterized with the transmission electron microscopy (TEM) images in Figure 1b–e. The initial Au NCs synthesized by a seed-mediated method exhibit regular cube shape have an edge length of about 45 nm. Figure 1c clearly confirms forming Au/Se Janus-like nanostructure, the obtained Se layer is ~26 nm. As for the as-synthesized Au/Bi₂Se₃ nanoparticles in Figure 1d, the average thicknesses of corresponding semiconductor shells are ~27 nm, and some irregular pores are distributed in them. After the overgrowth of Cu_2−xSe shells, the semiconductor shell’s thickness of Au/Bi₂Se₃/Cu_2−xSe nanoparticle increases to 36 nm. As we can see in Figure 1e, many semiconductor nanocrystals are attached onto the surface of hybrid nanoparticle, suggesting the formation of Cu_2−xSe nanocrystals on Bi₂Se₃ nanoshell.

The evolution of species and crystal structural for the products obtained at different reaction stages is monitored with corresponding X-ray diffraction (XRD) spectra and high-resolution TEM (HRTEM) images. As shown in Figure 2a, XRD measurement can be used to identify the lattice planes and crystal structure of synthesized Au/Bi₂Se₃ and Au/Bi₂Se₃/Cu_2−xSe. The data reveal that all the samples have Au phase, and four diffraction peaks at 38.18°, 44.39°, 64.57° and 77.54° are ascribed to (111), (200), (220) and (311) planes of cubic Au according to joint committee on powder diffraction standards (JCPDS) cards no. 04-0784. As for Au/Bi₂Se₃ nanoparticles, in addition to diffraction peaks of Au phases, two diffraction peaks at 29.45° and 43.69° belong to (015) and (110) planes of rhombohedral Bi₂Se₃ (JCPDS card no. 12-0732). The sample Au/Bi₂Se₃/Cu_2−xSe has three phases of Au, Bi₂Se₃ and Cu_2−xSe, owing to the formation of Cu_2−xSe nanocrystals. Besides the diffraction peaks of Au and Bi₂Se₃, four distinctive diffraction peaks appeared at 26.75°, 44.60°, 52.91° and 64.98° can be attributed to (111), (220), (311) and (400) planes of cubic Cu_2−xSe (JCPDS card no. 06-0680), respectively. In Figure 2b,c, the representative HRTEM images of individual hybrid nanoparticles are further employed to reveal the overgrowth of Cu_2−xSe on Au/Bi₂Se₃ nanoparticles. As we can see, two regions with different contrast can be clearly observed because of the different electron densities of metals and semiconductors, corresponding to Au core and semiconductor shell, respectively. Figure 2b shows the HRTEM image of individual Au/Bi₂Se₃ nanoparticle at its interface between metal and semiconductor, and the corresponding micro-area EDS result shown in Figure S2 suggests the presence of Au, Bi and Se elements. Uniform lattice fringe distribution is observed in the shell region, the lattice plane spacing of 0.30 nm can be referred to as the (015) lattice planes of rhombohedral Bi₂Se₃. The HRTEM image at the heterojunction of individual Au/Bi₂Se₃/Cu_2−xSe nanoparticle is displayed in Figure 2c. The lattice spacing of 0.30 nm and 0.33 nm can be indexed as the (015) crystal plane of rhombohedral Bi₂Se₃ and the (111) crystal plane of cubic Cu_2−xSe, respectively. All the results further confirm the overgrowth of Cu_2−xSe on Au/Bi₂Se₃ nanoparticle.

In addition, X-ray photoelectron spectroscopy (XPS) was carried out to analyze the chemical composition and binding states of the sample. Figure 3a shows XPS survey spectrum of the sample Au/Bi₂Se₃/Cu_2−xSe nanoparticles and Au, Se, Bi and Cu can be clearly observed. The corresponding high resolution XPS spectra of Se 3d, Cu 2p and Bi 4p are presented in Figure 3b–d. All XPS peaks were calibrated by C 1s peak at 284.8 eV. As we can see, the Se 3d spectrum in Figure 3b is deconvoluted into a doublet in a 3:2 area ratio, two XPS peaks appeared at 54.2 eV and 55.1 eV are assigned to Se 3d5/2 and Se 3d3/2, respectively, indicating the presence of Se²⁻ species. In Figure 3c, the binding energies at 158.5 eV and 163.8 eV with a splitting of 5.3 eV are ascribed to Bi 4f7/2 and Bi 4f5/2 peaks of Bi₂Se₃ shells, respectively, which are in agreement with the reported values [45]. The two peaks located at 932.9 eV and 952.9 eV are ascribed to Cu 2p3/2 and Cu 2p1/2, respectively. A weak satellite peak around 942.5 eV lies between these two peaks, indicating the main valence sate of cooper element is Cu⁺ [26]. The intensity of corresponding XPS peaks enhances with the increasing of Cu²⁺ concentration in Figure S3, suggesting an increase of Cu_2−xSe content. The results are consistent with TEM characterizations in Figure S1. In addition, a slightly positive shift (0.3 eV) of Bi 4f7/2 and Bi 4f5/2 in the Au/Bi₂Se₃/Cu_2−xSe can be observed in Figure 3c. The positive shift of Bi 4f indicates a decrease of the electron density near the Bi element caused by electrons transfer at the interface, forming an internal electric field directing from Bi₂Se₃ to Cu_2−xSe.

Figure 4a shows the evolution of absorption spectra of the samples in the different experimental stages. The absorption spectra of plasmonic metal nanocrystals are often strongly correlated to their morphologies and surrounding environments. The original Au NCs displays a strong resonance absorption band centered at 531 nm due to its surface plasmon resonance (SPR) behavior. After the formation of Janus-like Au/Se hetero nanostructures, the corresponding SPR peak shifts to 632 nm, attributing to the high refractive index of the amorphous Se shells. Subsequently, the SPR band of Au/Bi₂Se₃ hybrids further present an about 51 nm red shift. Meanwhile, a new absorption shoulder appears in the region of 420 to 450 nm, which can be ascribed to the formation of Bi₂Se₃ shells [46]. With the overgrowth of Cu_2−xSe on Au/Bi₂Se₃ hybrids, the resonance absorption band of the sample appears red shift and broaden, which may be attributed to the effective coupling between Au core and semiconductor shells. In addition, the influence of initial Cu²⁺ precursor concentration on the absorption spectra of Au/Bi₂Se₃/Cu_2−xSe hybrids was also presented in Figure 4b. All the obtained samples show the SPR peak centered at about 733 nm. As Cu²⁺ concentration increases from 0.05 M to 0.1 M, the corresponding resonance absorption peak becomes stronger and wider. Such a wide and strong spectral resonance absorption band matches well with the solar spectrum, allowing for their potential solar applications.

To evaluate the PEC water splitting performance of the obtained Au/Bi₂Se₃/Cu_2−xSe hybrids, pure Bi₂Se₃, Au/Bi₂Se₃ and Au/Bi₂Se₃/Cu_2−xSe NPs are loaded onto FTO working electrodes in a standard three-electrode electrochemical system, respectively. For comparison purposes, all samples have the same mass of corresponding components. Typical linear sweep voltammograms curves for five parallel electrodes are shown in Figure 5a. Under the simulated solar light, the photocurrent densities of the samples increase with the increase in applied potential. The photocurrent density of Au/Bi₂Se₃ electrode is 0.138 mA/cm² at the applied potential of 1.20 V vs. the reversible hydrogen electrode (RHE), which is 1.58 times that of pure Bi₂Se₃ electrode (0.089 mA/cm² at 1.20 V vs. RHE). Impressively, the photocurrent densities of 0.1-Au/Bi₂Se₃/Cu_2−xSe, 0.4-Au/Bi₂Se₃/Cu_2−xSe and 0.8-Au/Bi₂Se₃/Cu_2−xSe electrodes are 0.375, 0.706 and 0.237 mA/cm² at the applied potential of 1.20 V vs. RHE, respectively. 0.1-Au/Bi₂Se₃/Cu_2−xSe photoanode presents the highest photocurrent density among the surveyed samples, which implies efficient photoelectron generation and interface transfer processes. In an illuminated three-electrode cell, the obtained current–voltage data can be useful to assess the solar conversion efficiency of photoelectrodes, which allows us to optimize and identify the champion nanostructure for high performance photoelectrochemical water splitting. The photoconversion efficiency (η) of samples can be calculated by using the following equation:

η = I (1.23 − U)/J_light

where I is the photocurrent density (mA/cm²), U is the applied potential vs. RHE and J_light is the irradiance intensity of incident light (100 mW/cm²). In Figure 5b, the maximum photoconversion efficiency of the samples increases in the order of Bi₂Se₃, Au/Bi₂Se₃ and x-Au/Bi₂Se₃/Cu_2−xSe electrodes. A significant enhancement of photoconversion efficiency can be observed for Au/Bi₂Se₃ electrode by the introduction of Au NPs, owing to the plasmon-enhanced light harvesting. With the increasing of x, the maximum photoconversion efficiency of x-Au/Bi₂Se₃/Cu_2−xSe electrode first increases and then decreases. 0.1-Au/Bi₂Se₃/Cu_2−xSe electrode shows the optimal value of 0.11% at 0.95 V vs. RHE, which is 6.87 and 5.50 times that of Bi₂Se₃ (0.016% at 0.90 V vs. RHE) and Au/Bi₂Se₃ (0.02% at 0.90 V vs. RHE), demonstrating an excellent PEC water splitting performance.

To understand the enhanced mechanism of photocurrent response, photocurrent tests of the as-prepared Bi₂Se₃, Au/Bi₂Se₃ and x-Au/Bi₂Se₃/Cu_2−xSe electrodes were performed at an applied potential of 1.23 V vs. RHE, respectively. As shown in Figure 6a, it shows the I-t curves for three parallel electrodes under white-light (AM 1.5 G, 100 mW/cm²). The photocurrents of all electrodes present a transient increase after illumination and a clear photo-switching behavior. The stable photocurrents of 0.122 and 0.181 mA/cm² were correspondingly obtained for Bi₂Se₃ and Au/Bi₂Se₃ electrodes. By contrast, all x-Au/Bi₂Se₃/Cu_2−xSe electrodes exhibit a significant photocurrent enhancement. More notably, the photocurrent density of 0.1-Au/Bi₂Se₃/Cu_2−xSe electrode is 0.741 mA/cm². Compared with Bi₂Se₃ and Au/Bi₂Se₃ electrodes, 0.1-Au/Bi₂Se₃/Cu_2−xSe electrode presents 5.07 and 3.09 times of photoresponse enhancement, respectively. Combined with absorption spectrum in Figure 4a, the observed photoresponse enhancement may be attributed to plasmon-induced absorption enhancement and hot-electron injection. However, when x increases from 0.05 to 0.4, the photocurrent response of x-Au/Bi₂Se₃/Cu_2−xSe electrodes do not continue to increase with the light absorption enhancement but increases first and then decreases. The results indicate that there is another mechanism to improve the photoelectrochemical properties of the samples besides the enhanced light absorption. During the testing process, time-dependent photocurrent, XRD spectra and corresponding photo images of the sample are further recorded in Figure S4. The data show that the photocurrent density and crystal structure of the Au/Bi₂Se₃/Cu_2−xSe electrode remains almost unchanged, indicating the structural stability for potential applications. Subsequently, electrochemical impedance spectroscopy (EIS) experiments were carried out at open circuit potential under white light illumination to get further insight into the charge transfer and recombination processes. As shown in Figure 6b, R_s denotes the contact resistances of the electrochemical device, CPE denotes the capacitance phase element, R_ct denotes the interfacial charge transfer resistance and the arc radius on Nyquist plot is primarily proportional to the charge transfer resistance R_ct. The values of R_ct are calculated to be 1229, 668 and 432 Ω for pure Bi₂Se₃, Au/Bi₂Se₃ and 0.1-Au/Bi₂Se₃/Cu_2−xSe electrodes, respectively. Au/Bi₂Se₃/Cu_2−xSe electrode shows the lowest arc radius, indicating a lower charge transfer resistance. These results can be attributed to the formation of Au/Bi₂Se₃ and Bi₂Se₃/Cu_2−xSe multi-interface heterojunctions improving charge carries separation and transfer. The electron lifetime (${\tau }_n$) is related to the characteristic frequency peaks $f_{max}$ in Bode phase plots diagrams, that is ${\tau }_n=1/2\pi f_{max}$ [47]. As shown in Figure 6c, the characteristic peak values for pure Bi₂Se₃, Au/Bi₂Se₃ and Au/Bi₂Se₃/Cu_2−xSe electrodes (below 100 Hz) are 96.42, 37.76 and 1.89 Hz, respectively. The corresponding ${\tau }_n$ values are 1.65, 4.21 and 86.03 ms in turn. Besides that, we further compare the photoelectron transport performance in interfaces of x-Au/Bi₂Se₃/Cu_2−xSe electrodes. As shown in Figure 6d, the radius of arch in the impedance spectra appears decrease first and then increase along with the x value increasing. The data indicates that the right-sized Cu_2−xSe in Au/Bi₂Se₃/Cu_2−xSe is non-negligible to their photoelectrochemical properties. The corresponding characteristic frequency peak values in Bode phase plots are 1.85 Hz, 2.73 Hz and 4.44 Hz for 0.05, 0.1 and 0.4-Au/Bi₂Se₃/Cu_2−xSe electrodes, respectively. The electrons lifetime in 0.1-Au/Bi₂Se₃/Cu_2−xSe electrode is the longest among them, which means the FTO substrate possesses more opportunities to collect electrons produced by the electrode. As thus, all the results suggest that Au/Bi₂Se₃/Cu_2−xSe can effectively inhibit the charge recombination for superior solar-driven water splitting performance.

In addition, the Mott–Schottky diagrams and absorption spectra of pure Bi₂Se₃ and Cu_2−xSe NPs prepared at the same conditions are displayed in Figure S5. The pure Bi₂Se₃ and Cu_2−xSe are presented with a positive and negative slope in the Mott–Schottky diagrams, implying that Bi₂Se₃ is a p-type semiconductor and Cu_2−xSe is an n-type semiconductor. From the x-intercepts of linear extrapolations, the flat-band potentials of Bi₂Se₃ and Cu_2−xSe can be determined to be −0.21 V vs. RHE and 2.32 V vs. RHE, respectively. Their band gap energies (E_g) can be estimated as 1.65 and 2.24 eV from the Tauc plots (in Figure S4), respectively. Based on the above results, the direct Z-scheme migration pathway of the photogenerated carriers for Au/Bi₂Se₃/Cu_2−xSe electrode can be depicted in Figure 7. Here, when p-type Cu_2−xSe encounters with n-type Bi₂Se₃, electrons on the Bi₂Se₃ will migrate to Cu_2−xSe to obtain a Fermi level equilibrium, leaving an internal electric field from Bi₂Se₃ to Cu_2−xSe at the interface. Upon light irradiation, photoelectrons at the interface will be driven into the Bi₂Se₃ by the internal electric field, accelerating the charge carrier separation according to the direct Z-scheme mechanism. Considering the Fermi level of Au and the Schottky barrier (≈0.9 eV) at the Au-Bi₂Se₃ interface, the plasmon-induced hot electrons (1.69 eV) in Au core have sufficient energy to surmount the Schottky barrier and injected into the conduction band of Bi₂Se₃ shell. Simultaneously, the SPR band of Au/Bi₂Se₃/Cu_2−xSe hybrids lies in the same spectral region as the Bi₂Se₃ absorption edge, which will induce a resonance coupling between Au with Bi₂Se₃ and effectively promote the harvesting of visible light of the photoelectrode. Furthermore, the hollow and porous nanoshells not only maintain the independence of the plasmon nanostructures, but also provide a channel for the connection between the Au core and the solution. Therefore, hot holes can be rapidly eliminated by the hole scavengers to recycle the photoreduction/oxidation reactions.

3. Materials and Methods

3.1. Chemicals and Materials

Chloroauric acid (HAuCl₄·H₂O, 99.99%), hexadecyltrimethylammonium bromide (CTAB, 99%), sodium borohydride (NaBH₄,96%), L-ascorbic acid (AA, 99.7%), glycine acid (Gly, 99.5), selenium dioxide (SeO₂, 99%), bismuth (III) nitrate pentahydrate (Bi(NO₃)₃·5H₂O), copper(II) sulfate pentahydrate (CuSO₄·5H₂O, 99%), sodium sulfite anhydrous (Na₂SO₃, 97%), sodium sulfide nonahydrate (Na₂S·9H₂O, 98%) was purchased from Sinopharm Chemical Reagent Co. Ltd. (Shanghai, China). All chemicals were used as received and without further purification. All aqueous solutions were freshly prepared in the water obtained by ultra-pure water system from ZOOMWO Co. Ltd. (Hunan, China).

3.2. Synthesis of the Au Nanocubes

Au NCs were prepared by using a seed-mediated growth method [48]. The seed solution for Au NCs growth was prepared as follows: 0.25 mL of chloroauric acid (5.0 mM) and 3.75 mL of CTAB (0.1 M) was added into a vial at room temperature. One-hundred-and-fifty microliters of fresh-prepare and ice-cold NaBH₄ (0.01 M) was rapidly injected with vigorous stirring. Subsequently, the mixture was kept on gently stirring for 1 h. To prepare the growth solution, 12.8 mL of CTAB (0.1 M) and 1.6 mL of HAuCl₄ (0.01 M) were added into a 100 mL round-bottom flask with 62.4 mL deionized water. After then, 7.6 mL of ascorbic acid (0.1 M) was injected to the flask with gentle mixing. The color of the mixed solution changed from gold to colorless. Finally, 40 μL of 10 times diluted Au seed solution was added to the round-bottom flask and mix it gently for 30 s. The solution was allowed to stand in the flask overnight at 37 °C without stirring.

3.3. Synthesis of Porous Au/Bi₂Se₃ Nanoparticles

First, 60 mL of CTAB-stabilized Au nanocubes was centrifuged at 8000 rpm for 10 min to synthesize Au/Se intermediates. Second, the supernatant was removed, and the obtained precipitate was redispersed into a test tube with 20 mL of CTAB (0.2 M). Then, 0.2 mL of SeO₂ (0.1 M) and 1 mL of AA (0.1 M) were added successively. The solution was stirred for 5 h at 40 °C water bath. To prepared porous Au/Bi₂Se₃ nanoparticles, the mixture was transferred to 60 °C oil bath subsequently. Twenty milliliters of glycine (0.2 M), 1 mL of AA (0.1M) and 1.6 mL of Bi (NO₃)₃ (0.01 M, dissolved by 0.35 M HNO₃) were rapidly injected to the reaction solution. The mixed solution was kept at 80 °C under vigorously stirring for 5 h. The final products were washed by deionized water and collected by centrifugation at 8000 rpm for 10 min.

3.4. Synthesis of the Porous Au/Bi₂Se₃/Cu_2−xSe Nanoparticles

The obtained Au/Bi₂Se₃ nanoparticles were dispersed into the mixed solvents of 40 mL of CTAB (0.1 M) and 20 mL glycine (0.2 M). Next, a freshly prepared mixture of CuSO₄ (0.1 M) and AA (0.1 M) aqueous solution (the volume ratio of CuSO₄/AA was 1:4) was added under vigorous stirring. The mixture was allowed to react under vigorous stirring at 60 °C for 90 min. The obtained products were centrifuged at 8000 rpm for 9 min, and then redispersed in ultra-pure water for further use.

3.5. Photoelectrochemical (PEC) Tests

The PEC measurements were performed using a standard three-electrode potentiostat system on Instruments CS350 electrochemical workstation (Corrtest, Wuhan, China) with a working electrode, Pt counterelectrode and Ag/AgCl reference electrode. FTO substrate (1 cm × 2 cm) was first cleaned with ethanol and then water. In order to enhance the photoelectron collecting ability and stability of the electrode, the surface of FTO was modified by conductive silver adhesives. Then, the working electrode was prepared by depositing active films of the sample on the modified FTO substrate. The surface area of the sample exposed to the electrolyte was fixed at 1 cm². An aqueous solution containing 0.2 M of Na₂S and 0.2 M of Na₂SO₃ (pH ≈ 12.4) was used as the electrolyte to maintain the stability of the electrodes. Linear Sweep Voltammetry (LSV) curves had a scan rate of 10 mV s⁻¹. Chronoamperometry was applied at the potential of 1.23 V vs. RHE. Electrochemical impedance spectroscopy (EIS) was measured under open circuit conditions (AC amplitude).

3.6. Morphology and Structure Characterizations

TEM and HRTEM observations were performed with a FEI Tecnai F20 transmission electron microscope operated at 200 kV (FEI Company, Hillsboro, OR, USA) The XRD analyses were performed on a Bruker D8-advance X-ray diffractometer (Bruker AXS, Billerica, Germany) with Cu kαirradiation (λ = 1.5406 Å). Absorption spectra of the samples were measured using a TU-1810 UV-Vis’s spectrophotometer (Purkinje General Instrument Co. Ltd., Beijing, China). XPS (Thermo Fisher Scientific, USA) was recorded by a Thermo Fisher Scientific Esca lab 250Xi X-ray photoelectron spectrometer with a monochromatic Al Ka irradiation.

4. Conclusions

In summary, we have prepared Cu_2−xSe decorated Au/Bi₂Se₃ porous nanoparticles by a simple three-step method. Bi₂Se₃/Cu_2−xSe z-scheme heterojunction was introduced onto Au nanocube through a cation exchange method. The obtained Au/Bi₂Se₃/Cu_2−xSe hybrids possessed a broad and strong plasmon resonance absorption in the visible bands, owing to the effective coupling between Au core and semiconductor nanoshells. More intriguingly, the carrier lifetime of Au/Bi₂Se₃/Cu_2−xSe hybrid photoelectrodes could be further tailored with corresponding Cu_2−xSe content. Through parameter optimization, 0.1-Au/Bi₂Se₃/Cu_2−xSe hybrid photoanodes exhibited the longest electron lifetime (86.03 ms) among all the survey samples and demonstrated a decent photoelectrochemical performance enhancement in the tests. Compared with the pure Bi₂Se₃ (0.016% at 0.90 V vs. RHE) and Au/Bi₂Se₃ (0.02% at 0.90 V vs. RHE) nanoparticles photoanodes, the maximum photoconversion efficiency of porous Au/Bi₂Se₃/Cu_2−xSe hybrid photoanodes increased by 5.87 and 4.50 times under simulated sunlight illumination, all the results could be attributed to the synergistic effect of Z-scheme heterojunction and plasmon resonance enhancement. All the results suggest that Au/Bi₂Se₃/Cu_2−xSe porous nanostructures combine eco-friendliness with excellent sunlight harvesting capability and effectively inhibiting the charge recombination, which provide a new opportunity for efficient solar-driven water splitting.