Electrostatic Self-Assembly of CdS Quantum Dots with Co9S8 Hollow Nanotubes for Enhanced Visible Light Photocatalytic H2 Production

CdS quantum dots (CdS QDs) are regarded as a promising photocatalyst due to their remarkable response to visible light and suitable placement of conduction bands and valence bands. However, the problem of photocorrosion severely restricts their application. Herein, the CdS QDs-Co9S8 hollow nanotube composite photocatalyst has been successfully prepared by loading Co9S8 nanotubes onto CdS QDs through an electrostatic self-assembly method. The experimental results show that the introduction of Co9S8 cocatalyst can form a stable structure with CdS QDs, and can effectively avoid the photocorrosion of CdS QDs. Compared with blank CdS QDs, the CdS QDs-Co9S8 composite exhibits obviously better photocatalytic hydrogen evolution performance. In particular, CdS QDs loaded with 30% Co9S8 (CdS QDs-30%Co9S8) demonstrate the best photocatalytic performance, and the H2 production rate reaches 9642.7 μmol·g−1·h−1, which is 60.3 times that of the blank CdS QDs. A series of characterizations confirm that the growth of CdS QDs on Co9S8 nanotubes effectively facilitates the separation and migration of photogenerated carriers, thereby improving the photocatalytic hydrogen production properties of the composite. We expect that this work will facilitate the rational design of CdS-based photocatalysts, thereby enabling the development of more low-cost, high-efficiency and high-stability composites for photocatalysis.


Introduction
With the increasingly serious environmental pollution and the increasing demand for energy, the development and utilization of sustainable clean energy to achieve green development has become a hot topic [1][2][3][4].In recent years, photocatalytic hydrogen evolution has attracted much attention due to its advantages of zero carbon emission, high efficiency and sustainability, and is considered a promising energy conversion method [5][6][7].Therefore, the utilization of photocatalytic technology to produce hydrogen energy represents a feasible strategy for alleviating environmental pollution and energy crises [8][9][10].The practical application of photocatalytic hydrogen production technology is contingent upon three key factors: low cost, high efficiency and high stability [11,12].One of the most commonly employed modification strategies to improve the photocatalytic H 2 evolution properties of semiconductors is the introduction of precious metals (such as Au, Ag, Pd and Pt) through doping.Nevertheless, precious metals are limited and expensive.Consequently, the development of cost-effective, environmentally friendly and highly active photocatalysts represents a significant and pressing challenge [13,14].
In recent years, metal sulfides have become a research focus in the field of photocatalytic hydrogen evolution on account of their exceptional light absorption properties and unique electronic structure.Among these, CdS has been extensively studied owing to its appropriate band gap and position of energy bands [15,16].Moreover, CdS exhibits diverse morphologies and structures, involving zero-dimensional (0D) quantum dots, one-dimensional (1D) nanorods, two-dimensional (2D) nanosheets and three-dimensional (3D) cubes [17].CdS QDs are considered to be a promising photocatalytic material due to their small size (<10 nm), high electron mobility and abundant recoverable light [18,19].However, the issue of easy hole oxidation decomposition (photocorrosion) severely restricts the application of CdS [20].Among various strategies to alleviate CdS photocorrosion, the rational incorporation of a cocatalyst is an effective approach [21].Co 9 S 8 serves as a widely used cocatalyst with advantages such as easy availability, abundant active sites and adjustable chemical composition [22].In particular, the hollow-structured Co 9 S 8 possesses a large specific surface area and enhances the absorption of light by multiple reflections, which is of significant importance in improving photocatalytic properties.Additionally, the electrostatic self-assembly method is an efficient and environmentally friendly preparation method for nanoparticles, which is expected to prepare highly active photocatalysts [23].
Herein, the CdS QDs-Co 9 S 8 composite photocatalyst is successfully prepared through electrostatic self-assembly.Compared to blank CdS QDs, the CdS QDs-Co 9 S 8 composite demonstrates enhanced photocatalytic H 2 production performance.Notably, the optimal CdS QDs-30%Co 9 S 8 exhibits a photocatalytic hydrogen production rate of 9642.7 µmol•g −1 •h −1 , approximately 60.3 times that of blank CdS QDs.Cyclic experiments indicate that the introduction of Co 9 S 8 cocatalyst effectively prevents photocorrosion on the surface of CdS QDs.Moreover, subsequent characterizations confirm that loading Co 9 S 8 cocatalyst effectively promotes the separation and migration of photogenerated carriers, thereby improving the photocatalytic properties of CdS QDs.This work illustrates the significant role of Co 9 S 8 as a cocatalyst in the field of photocatalytic H 2 production, and is expected to provide a useful reference for the development of effective and stable photocatalysts.

Results and Discussion
The synthesis process of the CdS QDs-Co 9 S 8 composite photocatalyst is shown in Figure 1.Initially, the Co 9 S 8 nanotubes are achieved through a two-step hydrothermal approach, followed by treatment with APTES to impart a positive charge.Subsequently, the treated Co 9 S 8 nanotubes are subjected to an electrostatic assembly process with CdS QDs, resulting in the formation of the CdS QDs-Co 9 S 8 composite photocatalyst.A diagram of the prepared samples diagram is depicted in Figure S1.As illustrated in Figure S1a,b, CdS QDs exhibit a yellow powder, while Co 9 S 8 nanotubes display a black powder.Upon assembly of CdS QDs and Co 9 S 8 nanotubes, the resulting CdS QDs-Co 9 S 8 composite appears yellowish-green (Figure S1c).
Figure 2a,b display the Zeta potentials of APTES-modified Co 9 S 8 and CdS QDs suspension dispersed in deionized water, respectively.It can be observed that the Zeta potentials of APTES-modified Co 9 S 8 and CdS QDs are 13.8 mV and −30 mV, respectively, which means that APTES-modified Co 9 S 8 is positively charged, while CdS QDs is negatively charged.This result provides a good basis for the assembly of the CdS QDs-Co 9 S 8 composite [24].Morphological and microstructural analyses of CdS QDs, Co9S8 and CdS QDs-30%Co9S8 were conducted using scanning electron microscopy (SEM) and transmission electron microscope (TEM).As shown in Figure 3a, the TEM image reveals that the diameter of CdS QDs is roughly 4 nm, consistent with previous literature [25].Furthermore, as displayed in Figure 3b, the high-resolution TEM (HRTEM) image exhibits a lattice spacing of 0.35 nm corresponding to the (111) crystal face of CdS QDs, indicating its successful preparation [25].Meanwhile, the TEM and HRTEM images of Co9S8 (Figure S2) demonstrate the successful synthesis of Co9S8 nanotubes.As depicted in Figure S2b, the 0.23 nm of lattice spacing corresponds to the (420) crystal face of Co9S8.Morphological and microstructural analyses of CdS QDs, Co9S8 and CdS QDs-30%Co9S8 were conducted using scanning electron microscopy (SEM) and transmission electron microscope (TEM).As shown in Figure 3a, the TEM image reveals that the diameter of CdS QDs is roughly 4 nm, consistent with previous literature [25].Furthermore, as displayed in Figure 3b, the high-resolution TEM (HRTEM) image exhibits a lattice spacing of 0.35 nm corresponding to the (111) crystal face of CdS QDs, indicating its successful preparation [25].Meanwhile, the TEM and HRTEM images of Co9S8 (Figure S2) demonstrate the successful synthesis of Co9S8 nanotubes.As depicted in Figure S2b, the 0.23 nm of lattice spacing corresponds to the (420) crystal face of Co9S8.Morphological and microstructural analyses of CdS QDs, Co 9 S 8 and CdS QDs-30%Co 9 S 8 were conducted using scanning electron microscopy (SEM) and transmission electron microscope (TEM).As shown in Figure 3a, the TEM image reveals that the diameter of CdS QDs is roughly 4 nm, consistent with previous literature [25].Furthermore, as displayed in Figure 3b, the high-resolution TEM (HRTEM) image exhibits a lattice spacing of 0.35 nm corresponding to the (111) crystal face of CdS QDs, indicating its successful preparation [25].Meanwhile, the TEM and HRTEM images of Co 9 S 8 (Figure S2) demonstrate the successful synthesis of Co 9 S 8 nanotubes.As depicted in Figure S2b, the 0.23 nm of lattice spacing corresponds to the (420) crystal face of Co 9 S 8 .Figure 3c illustrates a hollow nanotube structure with a diameter of approximately 200 nm for Co 9 S 8 .As exhibited in Figure 3d, CdS QDs-30%Co 9 S 8 inherits the hollow nanotube structure of Co 9 S 8 .It is worth noting that the hollow structure exposes a large specific surface area and enhances the absorption of light by multiple reflections, which is of significant importance in improving the photocatalytic properties.Furthermore, it can be observed that CdS QDs are evenly decentralized on the Co 9 S 8 nanotubes.As presented in Figure 3e, the EDS spectra illustrate the presence of Co, Cd and S elements in the CdS QDs-30%Co 9 S 8 composite.Moreover, the composition of all the composite photocatalysts is quantitatively analyzed using inductively coupled plasma emission spectrometry (ICP-OES).As indicated in Table S1, as the Co 9 S 8 load increases, the proportion of the Co element rises, while the proportion of the Cd element decreases, consistent with the anticipated results.In addition, the element mapping results of CdS QDs-30%Co 9 S 8 indicate that CdS QDs are uniformly distributed on the surface of Co 9 S 8 nanotubes (Figure 3f).These results demonstrate the successful synthesis of the CdS QDs-30%Co 9 S 8 composite.
illustrates a hollow nanotube structure with a diameter of approximately 200 nm for Co As exhibited in Figure 3d, CdS QDs-30%Co9S8 inherits the hollow nanotube structure Co9S8.It is worth noting that the hollow structure exposes a large specific surface area a enhances the absorption of light by multiple reflections, which is of significant importan in improving the photocatalytic properties.Furthermore, it can be observed that CdS Q are evenly decentralized on the Co9S8 nanotubes.As presented in Figure 3e, the E spectra illustrate the presence of Co, Cd and S elements in the CdS QDs-30%Co composite.Moreover, the composition of all the composite photocatalysts is quantitativ analyzed using inductively coupled plasma emission spectrometry (ICP-OES).indicated in Table S1, as the Co9S8 load increases, the proportion of the Co element ris while the proportion of the Cd element decreases, consistent with the anticipated resu In addition, the element mapping results of CdS QDs-30%Co9S8 indicate that CdS QDs uniformly distributed on the surface of Co9S8 nanotubes (Figure 3f).These resu demonstrate the successful synthesis of the CdS QDs-30%Co9S8 composite.The crystal structure and phase composition of the prepared samples were investigated by X-ray diffraction (XRD).Figure 4a illustrates the XRD patterns of CdS QDs, Co 9 S 8 and CdS QDs-30%Co 9 S 8 .It can be observed that the XRD peak of CdS exhibits a relatively strong intensity, indicating its robust crystal phase.In contrast, the XRD peak of Co 9 S 8 displays a relatively weak intensity, suggesting its inferior crystal phase.For Co 9 S 8 , the diffraction peaks at 2θ = 29.9• , 31.4 • , 37.4 • , 39.5 • , 47.5 • , 52.3 • and 54.6 • correspond to the crystal planes (311), ( 222), (400), ( 331), ( 511), ( 400) and (531) of Co 9 S 8 , respectively (JCPDS: 65-1765) [26].As for CdS QDs, the characteristic peaks at 26.2 • , 43.6 • and 51.7 • can be related to the crystal faces (111), ( 220) and (311) of CdS (JCPDS: 75-1546), respectively [27].In addition, the XRD diffraction curve of CdS QDs-30%Co 9 S 8 is highly similar to that of CdS QDs, except that a weak peak at 39.5 • belongs to the (331) crystal plane of Co 9 S 8 , demonstrating the successful assembly of the CdS QDs-Co 9 S 8 composite.Furthermore, the (111) crystal face of CdS QDs exhibits a strong characteristic diffraction peak, resulting in a peak of Co 9 S 8 at 29.9 • masked by CdS QDs.The optical properties of a series of samples are determined by UV-vis diffuse reflectance spectroscopy (DRS).Figure 4b  investigated by X-ray diffraction (XRD).Figure 4a illustrates the XRD patterns of CdS QDs, Co9S8 and CdS QDs-30%Co9S8.It can be observed that the XRD peak of CdS exhibits a relatively strong intensity, indicating its robust crystal phase.In contrast, the XRD peak of Co9S8 displays a relatively weak intensity, suggesting its inferior crystal phase.For Co9S8, the diffraction peaks at 2θ = 29.9°,31.4°,37.4°, 39.5°, 47.5°, 52.3° and 54.6° correspond to the crystal planes (311), ( 222), (400), ( 331), ( 511), ( 400) and (531) of Co9S8, respectively (JCPDS: 65-1765) [26].As for CdS QDs, the characteristic peaks at 26.2°, 43.6° and 51.7° can be related to the crystal faces (111), ( 220) and (311) of CdS (JCPDS: 75-1546), respectively [27].In addition, the XRD diffraction curve of CdS QDs-30%Co9S8 is highly similar to that of CdS QDs, except that a weak peak at 39.5° belongs to the (331) crystal plane of Co9S8, demonstrating the successful assembly of the CdS QDs-Co9S8 composite.Furthermore, the (111) crystal face of CdS QDs exhibits a strong characteristic diffraction peak, resulting in a peak of Co9S8 at 29.9° masked by CdS QDs.The optical properties of a series of samples are determined by UV-vis diffuse reflectance spectroscopy (DRS).X-ray photoelectron spectroscopy (XPS) analysis (Figure 5) of the CdS QDs-30%Co9S8 composite is performed in order to further determine the chemical state and elemental composition of the prepared sample.As the survey spectra shown in Figure 5a, Co, Cd and S elements are present in the CdS QDs-30%Co9S8 composite, which further confirms the successful assembly of CdS QDs and Co9S8 cocatalyst.In the XPS spectra of Cd 3d (Figure 5b), the two characteristic peaks at 410.2 eV and 403.4 eV belong to Cd 3d3/2 and Cd 3d5/2, respectively, which demonstrates Cd exists in the form of +2 valence in the binary composite photocatalyst CdS QDs-30%Co9S8 [21].As illustrated in Figure 5c, the distinct peaks at the binding energies of 160.1 eV and 161.9 eV belong to S 2p3/2 and S 2p1/2, respectively, confirming the existence of S 2− [28].In addition, the XPS spectra of Co 2p displayed in Figure 5d can be divided into two spin-orbital dual peaks and two satellite peaks (identified as "Sat.").The first dual peaks at 780.3 eV and 776.6 eV and the second dual peaks at 796.8 eV and 794.5 eV can be attributed to Co 2p3/2 and Co 2p1/2, respectively, demonstrating the existence of Co 2+ and Co 3+ [29].The XPS results confirm that the X-ray photoelectron spectroscopy (XPS) analysis (Figure 5) of the CdS QDs-30%Co 9 S 8 composite is performed in order to further determine the chemical state and elemental composition of the prepared sample.As the survey spectra shown in Figure 5a, Co, Cd and S elements are present in the CdS QDs-30%Co 9 S 8 composite, which further confirms the successful assembly of CdS QDs and Co 9 S 8 cocatalyst.In the XPS spectra of Cd 3d (Figure 5b), the two characteristic peaks at 410.2 eV and 403.4 eV belong to Cd 3d 3/2 and Cd 3d 5/2 , respectively, which demonstrates Cd exists in the form of +2 valence in the binary composite photocatalyst CdS QDs-30%Co 9 S 8 [21].As illustrated in Figure 5c, the distinct peaks at the binding energies of 160.1 eV and 161.9 eV belong to S 2p 3/2 and S 2p 1/2 , respectively, confirming the existence of S 2− [28].In addition, the XPS spectra of Co 2p displayed in Figure 5d can be divided into two spin-orbital dual peaks and two satellite peaks (identified as "Sat.").The first dual peaks at 780.3 eV and 776.6 eV and the second dual peaks at 796.8 eV and 794.5 eV can be attributed to Co 2p 3/2 and Co 2p 1/2 , respectively, demonstrating the existence of Co 2+ and Co 3+ [29].The XPS results confirm that the prepared composite contains CdS and Co 9 S 8 , which indicates the successful preparation of this hybrid.
prepared composite contains CdS and Co9S8, which indicates the successful preparation of this hybrid.In order to compare the photocatalytic performance of pure CdS QDs and CdS QDs-Co9S8 composites illuminated by visible light, a photocatalytic hydrogen evolution experiment is conducted using TEOA as a sacrificial agent.As displayed in Figure 6a, on account of the serious recombination of photogenerated carriers, the blank CdS QDs exhibit low photocatalytic activity, which demonstrates a hydrogen production rate of 159.8 µmol•g −1 •h −1 .When CdS QDs are combined with 5%, 10%, 30% and 50% Co9S8 nanotubes, the different proportions of the CdS QDs-Co9S8 composites show enhanced photocatalytic activity.As the loading capacity of Co9S8 is increased, the photocatalytic H2 production rate of the CdS QDs-Co9S8 composites exhibits a corresponding increase.In particular, the optimal CdS QDs-30%Co9S8 composite photocatalyst demonstrated a hydrogen production rate of 9642.7 µmol•g −1 •h −1 , which is 60.3 times that of pure CdS QDs.Nevertheless, when the Co9S8 cocatalyst content continually increased, the hydrogen production rate of the CdS QDs-Co9S8 composite decreased.This phenomenon may be attributed to the high proportion of cocatalysts, which results in the masking of the CdS QDs' active sites during hydrogen evolution.As demonstrated in Table 1, the photocatalytic H2 evolution rate of the CdS QDs-30%Co9S8 composite is superior to that of similar photocatalysts documented in the literature.Furthermore, the photocatalytic stability of the CdS QDs-30%Co9S8 composite photocatalyst is evaluated by cyclic experiment.As illustrated in Figure 6b, the CdS QDs-30%Co9S8 composite photocatalyst demonstrates a stable photocatalytic activity following five cycles.These findings In order to compare the photocatalytic performance of pure CdS QDs and CdS QDs-Co 9 S 8 composites illuminated by visible light, a photocatalytic hydrogen evolution experiment is conducted using TEOA as a sacrificial agent.As displayed in Figure 6a, on account of the serious recombination of photogenerated carriers, the blank CdS QDs exhibit low photocatalytic activity, which demonstrates a hydrogen production rate of 159.8 µmol•g −1 •h −1 .When CdS QDs are combined with 5%, 10%, 30% and 50% Co 9 S 8 nanotubes, the different proportions of the CdS QDs-Co 9 S 8 composites show enhanced photocatalytic activity.As the loading capacity of Co 9 S 8 is increased, the photocatalytic H 2 production rate of the CdS QDs-Co 9 S 8 composites exhibits a corresponding increase.In particular, the optimal CdS QDs-30%Co 9 S 8 composite photocatalyst demonstrated a hydrogen production rate of 9642.7 µmol•g −1 •h −1 , which is 60.3 times that of pure CdS QDs.Nevertheless, when the Co 9 S 8 cocatalyst content continually increased, the hydrogen production rate of the CdS QDs-Co 9 S 8 composite decreased.This phenomenon may be attributed to the high proportion of cocatalysts, which results in the masking of the CdS QDs' active sites during hydrogen evolution.As demonstrated in Table 1, the photocatalytic H 2 evolution rate of the CdS QDs-30%Co 9 S 8 composite is superior to that of similar photocatalysts documented in the literature.Furthermore, the photocatalytic stability of the CdS QDs-30%Co 9 S 8 composite photocatalyst is evaluated by cyclic experiment.As illustrated in Figure 6b, the CdS QDs-30%Co 9 S 8 composite photocatalyst demonstrates a stable photocatalytic activity following five cycles.These findings demonstrate that the CdS QDs-Co 9 S 8 composite is an efficacious and stable photocatalyst.In addition, Figure 6c,d demonstrate that there is no obvious change in the SEM image and XRD pattern of the CdS QDs-30%Co 9 S 8 composites following cycling, which further shows that the composites have excellent stability.
demonstrate that the CdS QDs-Co9S8 composite is an efficacious and stable photocatalyst.In addition, Figure 6c,d demonstrate that there is no obvious change in the SEM image and XRD pattern of the CdS QDs-30%Co9S8 composites following cycling, which further shows that the composites have excellent stability.

Photocatalysts Light Sources Sacrificial Agents H2 (μmol•g −1 •h −1 ) Reference
CdS QDs-30% Co9S8 300 W Xe lamp (λ ≥ 420 nm) TEOA 9642.7 this work CdS/TiO2@Ti3C2 300 W Xe lamp (λ ≥ 420 nm) TEOA 3115.0 [30] CdS QDs/Ni2P/B-TiO2 300 W Xe arc lamp Na2S/Na2SO3 3303.9 [31] CdS/Au/KTaO3 Xe lamp (λ ≥ 420 nm) Na2S/Na2SO3 2892.0 [32] CdS QDs/CeO2 300 W Xe lamp (λ ≥ 300 nm) Na2S/Na2SO3 101.1 [33] Ni@NiO/CdS 500 W Xe lamp TEOA 4380.0 [34] CuS/CdS 300 W Xe lamp (λ ≥ 420 nm) lactic acid (10 vol%) 5617.0 [35] UiO-66-NH2@CdS 300 W Xe lamp (λ ≥ 420 nm) Na2S/Na2SO3 2028.5 [36]   The separation efficiency of photogenerated carriers can be evaluated through the photoluminescence (PL) measurement.As displayed in Figure 7a, the CdS QDs-30%Co 9 S 8 composite photocatalyst exhibits lower PL intensity than blank CdS QDs, indicating that the introduction of the Co 9 S 8 cocatalyst has an effective inhibition effect on the photogenerated electron-hole pair recombination, which can enhance the photocatalytic hydrogen evolution performance [39][40][41].Figure 7b shows the instantaneous photocurrent response of blank CdS QDs and CdS QDs-30%Co 9 S 8 [42].As demonstrated in Figure 7b, the photocurrent response of CdS QDs-30%Co 9 S 8 composite is significantly higher than that of blank CdS QDs, which indicates the improved photogenerated carrier separation efficiency of the CdS QDs-30%Co 9 S 8 composite [43][44][45].Additionally, the charge migration behavior at the catalyst-electrolyte interface is investigated through electrochemical impedance spectroscopy (EIS).Generally, a smaller radius of curvature results in lower resistance for the catalyst during the charge transfer process.As we can see from the EIS Nyquist diagram (Figure 7c), the CdS QDs-30%Co 9 S 8 composite exhibits a smaller curvature radius than that of the CdS QDs, which demonstrates that the CdS QDs-30%Co 9 S 8 composite represents a faster-photogenerated carrier transfer rate and a lower charge transfer resistance [46][47][48][49].As displayed in Figure 7d,e, both CdS QDs and Co 9 S 8 exhibit type IV isotherms in their N 2 adsorption-desorption isotherms, indicating the mesoporous nature of these materials.The pore size distribution of CdS QDs, Co 9 S 8 and CdS QDs-30%Co 9 S 8 is illustrated in Figure S3 and Table S2, further confirming their mesoporous characteristics.Furthermore, the BET surface areas of CdS QDs and Co 9 S 8 are 1.50 m 2 /g and 8.23 m 2 /g, respectively, while that of CdS QDs-30%Co 9 S 8 is 116.76 m 2 /g (Figure 7f).The larger BET surface area of the hybrid photocatalyst compared to that of Co 9 S 8 and CdS QDs suggests that the structure of quantum dots on hollow nanotubes can expand the catalyst's surface area, thereby enhancing the photocatalytic properties of the composite catalysts.Moreover, the N 2 adsorption isotherm and corresponding BET-specific surface areas of all other ratios of the composites have been investigated to identify discernible trends.As depicted in Figure S4, all composites exhibit higher BET-specific surface area than blank CdS and Co 9 S 8 , and their BET-specific surface area roughly decreases with increasing Co 9 S 8 load.This phenomenon may be attributed to the high loading amount of Co 9 S 8 , resulting in nanotube stacking and consequently reducing the composite's BET-specific surface area.
The band structure information of CdS and Co 9 S 8 can be acquired through UV-vis absorption spectra (Figure 4b) and Mott-Schottky plots (Figure 8).The band gap energy (E g ) of the synthesized samples is determined through the Tauc equation: (αhν) 2 = A(hν − E g ), where α, ν, h and A are the absorption coefficient, frequency of light, Planck's constant and proportionality constant, respectively.As depicted in Figure 8a,b, CdS QDs and Co 9 S 8 exhibit E g of 2.36 eV and 1.87 eV, respectively.The Mott-Schottky (M-S) method is employed to ascertain the semiconductor type and band potential.Figure 8c,d illustrates that both CdS and Co 9 S 8 display a positive slope, indicating their n-type semiconductor nature [50].From the M-S diagram, it is evident that the flat band potential (V fb ) for CdS QDs is −0.47 V, while that of Co 9 S 8 is −0.29 V (vs.Ag/AgCl).Since the conduction potential (E CB ) of n-type semiconductors is approximately −0.2 V negative than V fb , it can be calculated that the E CB of CdS and Co 9 S 8 are −0.67V and −0.49V (vs.Ag/AgCl), respectively [51].According to the conversion relationship, we determine that the E CB of CdS is −0.47 V, while that of Co 9 S 8 is −0.29 V (vs.NHE).Based on the E g of CdS and Co 9 S 8 , their valence band potential (E VB ) can be calculated as 1.89 V and 1.58 V using the formula E VB = E CB + E g .
Based on the aforementioned characterizations, a potential mechanism for visiblelight-driven photocatalytic hydrogen evolution by CdS QDs-Co 9 S 8 has been put forward.The band positions and band gaps of CdS and Co 9 S 8 have been determined through Mott-Schottky analysis and UV-vis DRS transformation plots.Since Co 9 S 8 (−0.29 V) exhibits a lower conduction band potential (E CB ) than CdS (−0.47 V), it suggests that the photogenerated electrons from CdS will be transferred to the CB of Co 9 S 8 .As exhibited in Figure 9, the irradiation of visible light results in the excited electrons in the valence band (VB) of CdS QDs jumping to the CB, accompanied by the generation of photogenerated holes in the VB.Due to the tight interfacial contact between CdS QDs and Co 9 S 8 , photogenerated electrons are transferred from the CB of CdS QDs to the CB of Co 9 S 8 instead of being trapped by holes.Subsequently, the electrons that have accumulated on Co 9 S 8 combine with H + to form H 2 .Meanwhile, the remaining photogenerated holes in the VB of CdS rapidly oxidize the sacrificial agent triethanolamine, forming a complete reaction cycle.Furthermore, the nanotube structure of Co 9 S 8 provides a multitude of active sites for photocatalytic hydrogen production reactions, and combined with the multiple reflections of light in the hollow structure of Co 9 S 8 , the photocatalytic H 2 evolution performance of the CdS QDs-Co 9 S 8 composite is significantly enhanced.The band structure information of CdS and Co9S8 can be acquired through UV-vis absorption spectra (Figure 4b) and Mott-Schottky plots (Figure 8).The band gap energy (Eg) of the synthesized samples is determined through the Tauc equation: (αhν) 2 = A(hν − Eg), where α, ν, h and A are the absorption coefficient, frequency of light, Planck's constant calculated that the ECB of CdS and Co9S8 are −0.67V and −0.49V (vs.Ag/AgCl), respectively [51].According to the conversion relationship, we determine that the ECB of CdS is −0.47 V, while that of Co9S8 is −0.29 V (vs.NHE).Based on the Eg of CdS and Co9S8, their valence band potential (EVB) can be calculated as 1.89 V and 1.58 V using the formula EVB = ECB + Eg.Based on the aforementioned characterizations, a potential mechanism for visiblelight-driven photocatalytic hydrogen evolution by CdS QDs-Co9S8 has been put forward.The band positions and band gaps of CdS and Co9S8 have been determined through Mott-Schottky analysis and UV-vis DRS transformation plots.Since Co9S8 (−0.29 V) exhibits a lower conduction band potential (ECB) than CdS (−0.47 V), it suggests that the photogenerated electrons from CdS will be transferred to the CB of Co9S8.As exhibited in Figure 9, the irradiation of visible light results in the excited electrons in the valence band (VB) of CdS QDs jumping to the CB, accompanied by the generation of photogenerated holes in the VB.Due to the tight interfacial contact between CdS QDs and Co9S8, photogenerated electrons are transferred from the CB of CdS QDs to the CB of Co9S8 instead of being trapped by holes.Subsequently, the electrons that have accumulated on Co9S8 combine with H + to form H2. Meanwhile, the remaining photogenerated holes in the VB of CdS rapidly oxidize the sacrificial agent triethanolamine, forming a complete reaction cycle.Furthermore, the nanotube structure of Co9S8 provides a multitude of active sites for photocatalytic hydrogen production reactions, and combined with the multiple reflections of light in the hollow structure of Co9S8, the photocatalytic H2 evolution performance of the CdS QDs-Co9S8 composite is significantly enhanced.

Preparation of CdS QDs
In a typical experiment, 1.7 mmol MPA (3-mercaptopropionic acid) and 1 mmol CdCl 2 were dissolved in 20 mL of deionized water.The pH was then modulated to about 10 through the addition of sodium hydroxide solution.The resulting solution was then diverted into a three-necked flask, which was sealed and the air outlet preserved.Subsequently, 5 mL of Na 2 S solution (0.2 mol/L) was added to the above solution in an atmosphere of argon gas and magnetically stirred.The solution was then heated to 373 K, after which the yellow solution was agitated for 0.5 h.Once the solution had cooled, 50 mL of ethanol was added to precipitate it.The resulting yellowish product was obtained after extraction, filtration, washing and drying.

Preparation of Co 9 S 8 Nanotubes
The preparation process of Co 9 S 8 nanotubes referred to the two-step hydrothermal method in previous work [52,53].Firstly, Co(CO 3 ) 0.35 Cl 0.20 (OH) 1.10 nanorods were synthesized as a precursor for Co 9 S 8 nanotubes.This was achieved by dissolving CoCl 2 •6H 2 O (5 mmol) and CH 4 N 2 O (5 mmol) in 40 mL deionized water and ultrasounding the solution for 30 min.Subsequently, the solution was diverted into a 50 mL Teflon autoclave and reacted in a 393 K oven for 10 h.The precipitate was then gathered through centrifugation and washed multiple times with anhydrous ethanol and deionized water.The pink precursor was obtained following drying at 333 K for several hours.Subsequently, the synthesized Co(CO 3 ) 0.35 Cl 0.20 (OH) 1.10 precursors (110 mg) were mixed to 40 mL of Na 2 S solution (5 mg/mL) in the Teflon liner and stirred for an hour.The liner was then diverted into a stainless-steel autoclave and heated to a temperature of 433 K for a period of 8 h.During the vulcanization process, the inner material of the rod-like precursor underwent a reaction and fell off, thereby obtaining the Co 9 S 8 of the hollow nanotube structure.Subsequently, the product was isolated through suction filtration, washed with anhydrous ethanol and deionized water and dried at 333 K for 12 h, and the dried product (black powder) was collected for further processing.

Positive Electrochemical Treatment of Co 9 S 8 Nanotubes
The prepared 100 mg Co 9 S 8 nanotubes were dispersed in 50 mL C 2 H 5 OH and ultrasonic until the solution was uniform.Then, 2 mL of APTES (3-aminopropyl triethoxysilane) solution was added to the ultrasonic-treated Co 9 S 8 nanotube ethanol solution and stirred for 20 min.Subsequently, the product was maintained in a water bath at 333 K for a period of four hours, centrifuged and washed with anhydrous ethanol and deionized water on several occasions.The obtained product was then dried in a 333 K oven and collected for use.

Electrostatic Assembly of CdS QDs-Co 9 S 8
Typically, 50 mg CdS QDs was dispersed in 50 mL deionized water and ultrasounded for 5 min.A certain proportion of 5%/10%/30% (2.5 mg/5 mg/15 mg) electropositive Co 9 S 8 nanotubes were dispersed in deionized water by the same method described above and ultrasonic.After ultrasound, the Co 9 S 8 nanotube solution was injected into the CdS QDs solution and stirred for a period of 2.5 h.Subsequently, the mixed solution was subjected to centrifugation and multiple washes with deionized water, after which it was dried in an oven at 333 K to yield the dried yellowish-green product.

Activity Evaluation of Photocatalytic H 2 Evolution
Photocatalytic H 2 production was conducted within a 50 mL closed quartz reactor.Typically, 1 mL of triethanolamine (TEOA) and 5 mL of deionized water were added to a sealed quartz reactor containing 5 mg of CdS QDs-Co 9 S 8 composite photocatalyst, followed by ultrasound until the solution was uniform.Subsequently, pure argon gas was implanted into the quartz reactor for half an hour to remove residuary air.A 300 W xenon lamp (PLS-SXE300D, Perfectlight, Beijing, China) with an ultraviolet cut-off filter (λ ≥ 420 nm) was used as the light source.Following a two-hour illumination period, 1 mL of mixed gas was injected into the gas chromatograph (GC7900, Techcomp, Shanghai, China) to detect the peak areas of hydrogen and argon, and the hydrogen production rate of the photocatalyst was then converted according to the hydrogen production coefficient given.Additionally, the stability of the CdS QDs-Co 9 S 8 composite photocatalyst was evaluated by conducting tests for 5 cycles under the same conditions after centrifugation, washing and drying.

Conclusions
In summary, Co 9 S 8 hollow nanotubes were prepared through a two-step hydrothermal approach as a cocatalyst, and the CdS QDs-Co 9 S 8 composite photocatalysts were successfully prepared through a straightforward electrostatic self-assembly method.The electrostatic self-assembly strategy ensures a tight interfacial contact between CdS QDs and Co 9 S 8 nanotubes.By adjusting the proportion of Co 9 S 8 nanotubes in the composite, the photocatalytic hydrogen evolution rate of the optimal CdS QDs-30%Co 9 S 8 nanotubes is 9642.7 µmol•g −1 •h −1 , approximately 60.3 times that of blank CdS QDs.The cyclic experiment demonstrates that the introduction of Co 9 S 8 cocatalysts effectively prevents photocorrosion on the surface of CdS QDs.A series of characterization experiments illustrate that the introduction of Co 9 S 8 hollow nanotubes resulted in a more uniform and dispersed growth of CdS QDs particles, as well as the promotion of the separation and migration of photogenerated carriers.As a result, the CdS QDs-Co 9 S 8 composite exhibits excellent activity and stability in photocatalytic hydrogen production.This work provides new perspectives for the rational construction of stable, environmentally friendly and highly active composite photocatalysts to realize efficient photocatalytic H 2 evolution.

Figure
Figure 2a and Figure 2b display the Zeta potentials of APTES-modified Co9S8 and CdS QDs suspension dispersed in deionized water, respectively.It can be observed that the Zeta potentials of APTES-modified Co9S8 and CdS QDs are 13.8 mV and −30 mV, respectively, which means that APTES-modified Co9S8 is positively charged, while CdS QDs is negatively charged.This result provides a good basis for the assembly of the CdS QDs-Co9S8 composite [24].

Figure
Figure 2a and Figure 2b display the Zeta potentials of APTES-modified Co9S8 and CdS QDs suspension dispersed in deionized water, respectively.It can be observed that the Zeta potentials of APTES-modified Co9S8 and CdS QDs are 13.8 mV and −30 mV, respectively, which means that APTES-modified Co9S8 is positively charged, while CdS QDs is negatively charged.This result provides a good basis for the assembly of the CdS QDs-Co9S8 composite [24].

Figure 2 .
Figure 2. Zeta potential of (a) APTES-modified Co 9 S 8 and (b) CdS QDs suspension dispersed in deionized water.
exhibits the light absorption curves of CdS QDs, Co 9 S 8 and the CdS QDs-30%Co 9 S 8 composite.The blank CdS presents a distinct absorption edge at near 570 nm.Moreover, Co 9 S 8 illustrates strong absorption across the entire spectral range, suggesting excellent light collection ability from ultraviolet to visible light regions.Notably, the CdS QDs-30%Co 9 S 8 composite displays superior light harvesting capability compared to CdS alone, which indicates the enhanced light absorption achieved through the introduction of the Co 9 S 8 cocatalyst in the composite photocatalyst.
Figure 4b exhibits the light absorption curves of CdS QDs, Co9S8 and the CdS QDs-30%Co9S8 composite.The blank CdS presents a distinct absorption edge at near 570 nm.Moreover, Co9S8 illustrates strong absorption across the entire spectral range, suggesting excellent light collection ability from ultraviolet to visible light regions.Notably, the CdS QDs-30%Co9S8 composite displays superior light harvesting capability compared to CdS alone, which indicates the enhanced light absorption achieved through the introduction of the Co9S8 cocatalyst in the composite photocatalyst.

Figure 6 .
Figure 6.(a) Photocatalytic hydrogen production rates of blank CdS QDs and CdS QDs-Co9S8 composite.(b) Cyclic stability test of CdS QDs-30%Co9S8 photocatalytic hydrogen production.(c) The SEM images of CdS QDs-30%Co9S8 composite after cyclic test.(d) XRD patterns of the CdS QDs-30%Co9S8 before and after cyclic test.

Figure 6 .
Figure 6.(a) Photocatalytic hydrogen production rates of blank CdS QDs and CdS QDs-Co 9 S 8 composite.(b) Cyclic stability test of CdS QDs-30%Co 9 S 8 photocatalytic hydrogen production.(c) The SEM images of CdS QDs-30%Co 9 S 8 composite after cyclic test.(d) XRD patterns of the CdS QDs-30%Co 9 S 8 before and after cyclic test.

Figure 9 .
Figure 9. Mechanism diagram of CdS QDs-Co 9 S 8 in photocatalytic hydrogen production driven by visible light.

Table 1 .
Contrast of the H2 production performance of the CdS-based photocatalysts.

Table 1 .
Contrast of the H 2 production performance of the CdS-based photocatalysts.