Constructing the Sulfur-Doped CdO@In2O3 Nanofibers Ternary Heterojunction for Efficient Photocatalytic Hydrogen Production

An S-doped CdO@In2O3 nanofiber was successfully designed by in-situ electrospinning along and subsequent calcination treatment. Under artificial sunlight illumination, the S/CdO@In2O3-25 displayed a superior photocatalytic hydrogen evolution rate of 4564.58 μmol·g−1·h−1, with approximately 22.0 and 1261.0-fold of those shown by the S/CdO and S/In2O3 samples, respectively. The experimental and theoretical analyses illustrate that the unique one-dimensional (1D) nanofiber morphology and rich oxygen vacancies optimized the electronic structure of the nanofibers and adsorption/desorption behaviors of reaction intermediates, contributing to the realization of the remarkable solar-to-H2 conversion efficiencies. Moreover, the staggered band structure and intimate contact heterointerfaces facilitate the formation of a type-II double charge-transfer pathway, promoting the spatial separation of photoexcited charge carriers. These results could inform the design of other advanced catalyst materials for photocatalytic reactions.


Introduction
The consumption of non-renewable energy has aggravated energy shortages and environmental pollution, largely hindering the progress of human civilization. It has been reported that the sustainable and renewable energy sources are key to relieving these issues [1,2]. Among these renewable energy sources, hydrogen has been considered an important energy carrier because of its pollution-free and high energy density characteristics; moreover, it is also expected to act as the most promising alternative fuel in the future [3,4]. Compared with other hydrogen generation technologies, photocatalytic water splitting induced by catalysts to generate hydrogen in large quantities has been considered as one of the most promising strategies to overcome the global energy crisis [5]. Therefore, the solar-to-hydrogen technique has attracted considerable attention. Currently, in order to attain preferable hydrogen evolution activity, the development of efficient and eco-friendly hydrogen photocatalyst materials, which is a challenging task, is necessary [6]. In recent decades, various oxide semiconductor materials for photocatalysis such as In 2 O 3 [7], ZnO [8], TiO 2 [9], CdO [10], and CdS [11], have been widely explored. As a typical semiconductor, low-cost CdO is an n-II-VI material with outstanding electrical conductivity and a direct band gap [12,13]. It has been widely used in solar cells [14], transparent electrodes [15], electrochemical capacitors, gas sensors, and as a photocatalyst [16]. In addition, In 2 O 3 is an important n-type semiconductor widely utilized in optoelectronic devices and photocatalysis because of its suitable band alignment, good photothermal Nanomaterials 2023, 13, 401 2 of 14 stability, low toxicity, and unique optical/electronic properties [17,18]. Unfortunately, these single photocatalyst materials suffer from rapid recombination of charge carriers and low light utilization rates, which severely hinders their further application [19]. It is well known that the performance of photocatalysts can be improved by morphological adjustment [20], functional interface engineering [21], and heterojunction construction [22], which can greatly overcome the above shortcomings [23,24]. Thus, various advanced heterostructure photocatalysts such as CdTe-Bi 2 S 3 [25], MoS 2 @CoMoS 4 [26], Bi 2 S 3 @CoO [27], BiOI/Ag/PANI [28], α-Fe 2 O 3 /CeO 2 [29], ZnO/ACN/MnO 2 [30], α-Fe 2 O 3 /g-C 3 N 4 [31], and Bi 2 WO 6 /TiO 2 [32] have been extensively developed.
In addition to the above-improved strategies, the doping of non-metallic atoms (such as N, C, and S), which could result in the formation of abundant vacancies that could serve as potential charge capture centers that promote the spatial separation of carriers, has also been effective [33,34]. Among the reported micro/nanomorphologies, one-dimensional (1D) nanomaterials exhibit attractive application merits [35]. In particular, nanofibers have the sizeable 1D morphology, highly aligned nanoparticles and a large pore structure, which could provide rich charge-transfer channels for effective photocatalytic reactions. Electrospinning has been proposed to construct various 1D nanofibers materials for photocatalytic hydrogen evolution and dye degradation [36][37][38][39][40].
In this paper, we report the synthesis of an ingenious S-doped CdO@In 2 O 3 hybrid nanofiber via in-situ electrospinning and facile calcination. As the S-doped CdO@In 2 O 3 nanofiber was used as a photocatalyst, clean water was effectively reduced to H 2 within a single integrated system under the simulated sunlight irradiation. The superior photocatalytic hydrogen evolution activity in the S-doped CdO@In 2 O 3 hybrid can be attributed to its improved sunlight absorption ability, suppressed photo-induced carrier recombination, and accelerated charge separation properties. As an encouraging result, the optimized hydrogen evolution rate of the as-obtained hybrid can reach 4564.58 µmol·g −1 ·h −1 , which exceeds that of the individual ones. To the best of our knowledge, this is the first study to report the synthesis of an S-doped CdO@In 2 O 3 ternary heterojunction for efficient H 2 evolution under simulated sunlight irradiation. The results herein are expected to be of interest to researchers.

Preparation of Samples
The electrospinning process is illustrated in Figure 1. Specifically, a certain amount of In(NO 3 ) 3 ·4.5H 2 O, Cd(NO 3 ) 2 ·4H 2 O and sulfourea with different dosages were dissolved in 2 mL DMF containing 0.8 g PVP and 8 mL ethanol. After stirring for 12 h, a mixed gel containing sulfourea, In(NO 3 ) 3 ·4.5H 2 O, Cd(NO 3 ) 2 ·4H 2 O, and PVP was obtained. The prepared precursor sol was then poured into the injector using metal needles. The actual distance between the tip of the needle and Al foil was approximately 17 cm. When a positive voltage of 1.5 kV and a negative pressure were applied, the jet was stretched by an electrostatic force to produce hybrid nanofibers. The as-prepared product was then dried in an oven for 12 h and calcined in a muffle furnace at 480 • C for 4 h. The as-obtained samples were denoted as S/CdO@In 2 O 3 , where the S/CdO@In 2 O 3 samples with molar ratios of 0.15, 0.25, or 0.35 of In 2 O 3 and CdO were denoted as S/CI-15, S/CI-25, and S/CI-35, respectively. For comparison, pure In 2 O 3 , CdO, CdO@In 2 O 3 -25 (CI-25), S/CdO (S/C), and S/In 2 O 3 (S/I) were also synthesized using the same procedure.
. Figure 1. The scheme diagram of electrospinning process of S/CdO@In2O3 hybrid nanofibers during the calcinations process.

Characterization of the Materials
X-ray photoelectron spectroscopy (XPS) (ESCALAB Xi+, Thermo Fisher Scientific, Waltham, MA, USA) was used to estimate the elemental surface chemical states. Specifically, the vacuum degree of the analysis chamber is 8 × l0 −10 Pa. The excitation source is Al Kα X-ray (hv = 1486.6 eV) with the working voltage of 12.5 kV and the filament current of 16 mA. The results are the signal accumulation of 10 cycles test. The pass energy of the survey spectrum and narrow spectrum is 100 and 30 eV, respectively, with the step size of 0.1 eV and the dwell time of 40-50 ms. The light spot diameter is about 650 μm. All the XPS spectra were calibrated by the 284.8 eV of C1s. Powder X-ray diffraction (XRD) patterns were measured in the 2θ region of 5-80° using a SmartLab SE (Rigaku, Tokyo, Japan) diffractometer with a copper target as the radiation source. Photoluminescence (PL) emission spectra were measured using a F-7000 fluorescence spectrometer (Hitachi, Tokyo, Japan). Fourier transform infrared (FT-IR) spectroscopy was performed using a IS10 (Nicolet, WI, USA) spectrometer. The morphological characteristics of the samples were observed using a Sigma 300 field-emission scanning electron microscope (FE-SEM) (Zeiss, Oberkochen, Germany) (FE-SEM). Transmission electron microscopy (TEM)/high resolution transmission electron microscopy (HR-TEM) was performed using an a Tecnai F20 microscope (FEI, Portland, OR, USA). The solid ultraviolet-visible (UV-vis) diffuse reflectance (UV-vis DRS) spectra were determined using a UV-2600 spectrometer (Shimadzu, Tokyo, Japan). Raman spectroscopy was performed on a Raman spectrometer (Renishaw, London, UK).

The Preparation of Catalyst Films and Photoelectrochemical Measurements
A three-electrode configuration with a quartz battery was used to evaluate the photoelectrochemical properties of the samples. A 0.1 mol/L Na2SO4 solution was used as the electrolyte, and the prepared catalyst films, platinum wire, and saturated calomel electrode were used as the working, counter, and reference electrodes, respectively. An electrochemical workstation (CHI650E) (Chenhua, Shanghai, China) was used. Specifically, the FTO glass was cleaned ultrasonically in acetone half a hour, rinsed with distilled water and ethanol, and dried at 60 °C. And 10.0 mg catalysts were dispersed in l mL absolute ethanol with 15 μL Nafion solution, and then ultrasound treated for 10 min. The suspension was evenly loading onto the FTO, glass, and then dried at 60 °C under vacuum conditions. The transient photocurrent response test was performed using a 400 W xenon lamp, and the stable time was 20 s for signal acquisition. The electrochemical impedance spectroscopy (EIS) (Chenhua, Shanghai, China) test frequency range was 100 KHz-0.01 Hz, and the initial voltage was measured for the open circuit voltage acquisition signal.

Characterization of the Materials
X-ray photoelectron spectroscopy (XPS) (ESCALAB Xi+, Thermo Fisher Scientific, Waltham, MA, USA) was used to estimate the elemental surface chemical states. Specifically, the vacuum degree of the analysis chamber is 8 × l0 −10 Pa. The excitation source is Al Kα X-ray (hv = 1486.6 eV) with the working voltage of 12.5 kV and the filament current of 16 mA. The results are the signal accumulation of 10 cycles test. The pass energy of the survey spectrum and narrow spectrum is 100 and 30 eV, respectively, with the step size of 0.1 eV and the dwell time of 40-50 ms. The light spot diameter is about 650 µm. All the XPS spectra were calibrated by the 284.8 eV of C1s. Powder X-ray diffraction (XRD) patterns were measured in the 2θ region of 5-80 • using a SmartLab SE (Rigaku, Tokyo, Japan) diffractometer with a copper target as the radiation source. Photoluminescence (PL) emission spectra were measured using a F-7000 fluorescence spectrometer (Hitachi, Tokyo, Japan). Fourier transform infrared (FT-IR) spectroscopy was performed using a IS10 (Nicolet, WI, USA) spectrometer. The morphological characteristics of the samples were observed using a Sigma 300 field-emission scanning electron microscope (FE-SEM) (Zeiss, Oberkochen, Germany) (FE-SEM). Transmission electron microscopy (TEM)/high resolution transmission electron microscopy (HR-TEM) was performed using an a Tecnai F20 microscope (FEI, Portland, OR, USA). The solid ultraviolet-visible (UV-vis) diffuse reflectance (UV-vis DRS) spectra were determined using a UV-2600 spectrometer (Shimadzu, Tokyo, Japan). Raman spectroscopy was performed on a Raman spectrometer (Renishaw, London, UK).

The Preparation of Catalyst Films and Photoelectrochemical Measurements
A three-electrode configuration with a quartz battery was used to evaluate the photoelectrochemical properties of the samples. A 0.1 mol/L Na 2 SO 4 solution was used as the electrolyte, and the prepared catalyst films, platinum wire, and saturated calomel electrode were used as the working, counter, and reference electrodes, respectively. An electrochemical workstation (CHI650E) (Chenhua, Shanghai, China) was used. Specifically, the FTO glass was cleaned ultrasonically in acetone half a hour, rinsed with distilled water and ethanol, and dried at 60 • C. And 10.0 mg catalysts were dispersed in l mL absolute ethanol with 15 µL Nafion solution, and then ultrasound treated for 10 min. The suspension was evenly loading onto the FTO, glass, and then dried at 60 • C under vacuum conditions. The transient photocurrent response test was performed using a 400 W xenon lamp, and the stable time was 20 s for signal acquisition. The electrochemical impedance spectroscopy (EIS) (Chenhua, Shanghai, China) test frequency range was 100 KHz-0.01 Hz, and the initial voltage was measured for the open circuit voltage acquisition signal. The Mott-Schottky curves were tested at 500 and 1500 Hz with the measured open-circuit voltage at the setting center −(0.5-1) V as the starting voltage, +(0.5-1) V as the termination voltage, and 0.01 V as the amplitude.

Photocatalytic Performance Measurement
Photocatalytic hydrogen production as performed in a cylindrically irradiated quartz vessel. Visible light (λ > 400 nm, 300 W Xe, CEL-HXF300) (Zhongjiao Jinyuan Technology Ltd, Beijing, China) was the light source for photocatalytic reactions. The light source was placed at a distance of 12 cm on the surface of the reaction solution, and all UV light with a wavelength less than 400 nm was removed using a 400 nm cutoff filter. The as-synthesized catalyst powder (0.05 g) was dispersed in a 50 mL solution containing 0.2 M Na 2 S and 0.5 M Na 2 SO 3 , and 800 µL of potassium chloroplatinate was added as a cocatalyst. Before the reaction, the system was bubbled with nitrogen for 20 min to eliminate the air inside the system and ensure that the entire system was under anaerobic conditions. Using high purity nitrogen as the carrier gas, the photocatalytic H 2 generation performance was determined by TCD gas chromatography (GC-7920) (Zhongjiao Jinyuan Technology Ltd, Beijing, China). Blank experiments were performed without a catalyst or light.

Theoretical Calculation
DFT calculations were performed in the Vienna ab initio simulation package (VASP). A spin-polarized GGA PBE functional [41], all-electron plane-wave basis sets with an energy cutoff of 520 eV, and a projector augmented wave (PAW) method were adopted [42,43]. A (3 × 3 × 1) Monkhorst-Pack mesh was used for the Brillouin-zone integrations to be sampled. The conjugate gradient algorithm was used in the optimization. The convergence threshold was set 1 × 10 −5 eV in total energy and 0.02 eV/Å in force on each atom. In the simulations, the non-periodic boundary condition is employed, and the molecular model of a orthorhombic In 2 O 3 (1 × 1 × 1) and hexagonal CdO (2 × 2 × 1) were established by using Materials Studio. The orthorhombic In 2 O 3 and hexagonal CdO composite material were formed by using Materials Studio. To find the thermal stable morphology and achieve a conformation with minimum potential energy, energy minimization was performed, and these minimum energy conformations were used as the initial status in the following electronic structure simulations [44] and Materials Studio software was used for visualization and plotting. The adsorption energy change (∆Eabs) was determined as follows: where Etotal is the total energy for the adsorption state, Eslab is the energy of pure surface, and Emol is the energy of adsorption molecule.
The free energy change (∆G) for adsorptions was determined as follows: where Etotal is the total energy for the adsorption state, E slab is the energy of pure surface, Emol is the energy of adsorption molecule, ∆E ZPE is the zero-point energy change, and ∆S is the entropy change. As the vibrational entropy of H* in the adsorbed state is small, the entropy of adsorption of 1/2 H 2 is S H ≈ −0.5S 0H2 , where S 0H2 is the entropy of H 2 in the gas phase at the standard conditions. Therefore, the overall corrections were taken as in ∆G H* = E tota l − E slab − E H2 /2 + 0.24 eV, where E H2 is the energy of H 2 in the gas phase.

Structures and Morphologies Characterization of S/CdO@In 2 O 3 Nanofibers
The phase structures of the synthesized catalysts were studied using XRD, as shown in Figure 2a [45]. In contrast, the diffraction peaks of the CdO samples were located at 33.0 • , 38.3 • , 55.3 • , 65.9 • and 69.2 • , and were indexed to the cubic phase (JCPDS No. 05-0640). The diffraction results show the peak of pure CdO is much sharper than that of individual In 2 O 3 , indicating that CdO has much better crystallinity. It can be observed that the prepared composites contain the characteristic diffraction peaks of both In 2 O 3 and CdO species. This implies that the asconstructed hybrid catalyst contains metal oxides of In 2 O 3 and CdO. After electrospinning with sulfourea, no variations were observed in the crystalline structure of CdO@In 2 O 3 , which may be attributed to its low S content. Infrared and Raman characterizations were carried out to study the chemical bonds and structural properties of the catalysts. Figure  and were indexed to the cubic phase (JCPDS No. 05-0640). The diffraction results show the peak of pure CdO is much sharper than that of individual In2O3, indicating that CdO has much better crystallinity. It can be observed that the prepared composites contain the characteristic diffraction peaks of both In2O3 and CdO species. This implies that the asconstructed hybrid catalyst contains metal oxides of In2O3 and CdO. After electrospinning with sulfourea, no variations were observed in the crystalline structure of CdO@In2O3, which may be attributed to its low S content. Infrared and Raman characterizations were carried out to study the chemical bonds and structural properties of the catalysts. Figure  2b represents the FTIR spectra of the samples. A strong band was observed at 3458 cm −  Moreover, the morphology of the samples was analyzed using SEM and TEM. As shown in Figure 3a, the surface of the S/CdO@In2O3-25 precursor nanofibers is smooth with a relatively uniform diameter (about 660 nm). After calcination, the fiber surface became rough, and its diameter showed a downward trend with a porous structure owing to the decomposition of the raw materials ( Figure S1). Furthermore, S/CdO@In2O3-25 Moreover, the morphology of the samples was analyzed using SEM and TEM. As shown in Figure 3a, the surface of the S/CdO@In 2 O 3 -25 precursor nanofibers is smooth with a relatively uniform diameter (about 660 nm). After calcination, the fiber surface became rough, and its diameter showed a downward trend with a porous structure owing to the decomposition of the raw materials ( Figure S1). Furthermore, S/CdO@In 2 O 3 -25 maintained a fibrous morphology with nanosphere-attached surfaces, as shown in Figure 3b  maintained a fibrous morphology with nanosphere-attached surfaces, as shown in Figure  3b. Further TEM observations reveal that the morphology of S/CdO@In2O3-25 after calcination consists of a shorter fiber length and coarser surface (Figure 3c). Figure 3d shows that the lattice fringes of In2O3 and CdO are 0.295 and 0.271 nm, respectively [40]. The oxygen vacancies are shown in Figure 3e, where some red dotted circles represent the vacancy sites. Further elemental mapping analysis and EDS spectra suggest that the In, Cd, S, and O elements are evenly distributed on the entire S/CdO@In2O3-25 nanofibers, as shown in Figures 3f-i and S2. The multiple components are closely combined, favoring the formation of ternary heterojunctions with close-contact interfaces. Furthermore, the electronic interaction and elemental chemical states of the as-prepared samples were examined using XPS. As shown in Figure S3, In, Cd, O, S elements are found in the composites, as expected, and no other impurities are detected, which agrees well with the EDS analysis results. This indicates that S/CdO@In2O3-25 hybrid nanofibers were successfully synthesized by in situ electrospinning. From the In 3d spectrum (Figure 4a), it is found that the binding energies (BEs) of pure In2O3 are 444.01 and 451.56 eV, which correspond to In 3d5/2 and In 3d3/2, respectively [47]. After coupling with CdO, the BEs become 444.12 and 451.67 eV. These two peaks also exist in the fine S/CdO@In2O3-25 spectra, with binding energies of 444. 19 and 451.71 eV. Both exhibit a positive shift relative to that of In2O3 alone. The positive shift for the In 3d species may be caused by the strong interaction among the S, In, and Cd elements, which ameliorates the interfacial charge-transfer behaviors. In Figure 4b, the Cd 3d spectrum of pure CdO exhibits two asymmetric peaks at 403.91 and 410.60 eV, which may belong to the spin-orbit splitting of Cd 3d5/2 and Cd 3d3/2, respectively [48], indicating that the chemical states of Cd in the nanocomposite are +2 [49]. After binding with In2O3, the BEs become 405.09 and 411.78 eV. The BEs in S/CdO@In2O3-25, in which positive shifts might give rise to lattice distortion, are 405.66 and 412.42 eV. Figure 4c shows that the O 1s of bare CdO has two asymmetric peaks at 528.34 and 531.49 eV, and the main oxygen peak at 531.49 eV confirms that the O 2-oxidation state exists in CdO [13]. The small oxygen peak at 528.34 eV is attributed to chemically adsorbed oxygen. Moreover, the O 1s spectrum of bare In2O3 also has two asymmetric peaks located at 529.58 and 531.46 eV. The lower-energy peak at 529.58 eV is attributed to the lattice oxygen of In2O3, and the higher-energy peak at 531.46 eV is caused by oxygen defects [50,51]. After the combination of CdO and In2O3, the two peaks for O 1s exhibit a positive shift and are located at 529.69 and 531.73 eV. When the samples are doped with sulfur, the positive shifts of the peaks increase and the two corresponding peaks are located at 530.53 and 532.83 eV. This indicates that S doping can increase oxygen vacancies, which may be conducive to the rapid separation of the photoinduced charge carriers. Furthermore, the S 2p spectrum (Figure 4d) can be divided into two peaks at 162.42 and 169.05 eV. These peaks are part of the spin orbits of S 2p3/2 and S 2p1/2 of S 2- [52,53], which have complementary oxygen atoms in the lattice of CdO and In2O3, thereby confirming the presence of oxygen vacancies. The BEs region between 166 and 172 eV can be ascribed to the oxidized sulfur species, and the peak located at  Furthermore, the electronic interaction and elemental chemical states of the as-prepared samples were examined using XPS. As shown in Figure S3, In, Cd, O, S elements are found in the composites, as expected, and no other impurities are detected, which agrees well with the EDS analysis results. This indicates that S/CdO@In 2 O 3 -25 hybrid nanofibers were successfully synthesized by in situ electrospinning. From the In 3d spectrum (Figure 4a), it is found that the binding energies (BEs) of pure In 2 O 3 are 444.01 and 451.56 eV, which correspond to In 3d 5/2 and In 3d 3/2 , respectively [47]. After coupling with CdO, the BEs become 444.12 and 451.67 eV. These two peaks also exist in the fine S/CdO@In 2 O 3 -25 spectra, with binding energies of 444.19 and 451.71 eV. Both exhibit a positive shift relative to that of In 2 O 3 alone. The positive shift for the In 3d species may be caused by the strong interaction among the S, In, and Cd elements, which ameliorates the interfacial charge-transfer behaviors. In Figure 4b, the Cd 3d spectrum of pure CdO exhibits two asymmetric peaks at 403.91 and 410.60 eV, which may belong to the spin-orbit splitting of Cd 3d 5/2 and Cd 3d 3/2 , respectively [48], indicating that the chemical states of Cd in the nanocomposite are +2 [49]. After binding with In 2 O 3 , the BEs become 405.09 and 411.78 eV. The BEs in S/CdO@In 2 O 3 -25, in which positive shifts might give rise to lattice distortion, are 405.66 and 412.42 eV. Figure 4c shows that the O 1s of bare CdO has two asymmetric peaks at 528.34 and 531.49 eV, and the main oxygen peak at 531.49 eV confirms that the O 2oxidation state exists in CdO [13]. The small oxygen peak at 528.34 eV is attributed to chemically adsorbed oxygen. Moreover, the O 1s spectrum of bare In 2 O 3 also has two asymmetric peaks located at 529.58 and 531.46 eV. The lower-energy peak at 529.58 eV is attributed to the lattice oxygen of In 2 O 3 , and the higher-energy peak at 531.46 eV is caused by oxygen defects [50,51]. After the combination of CdO and In 2 O 3 , the two peaks for O 1s exhibit a positive shift and are located at 529.69 and 531.73 eV. When the samples are doped with sulfur, the positive shifts of the peaks increase and the two corresponding peaks are located at 530.53 and 532.83 eV. This indicates that S doping can increase oxygen vacancies, which may be conducive to the rapid separation of the photoinduced charge carriers. Furthermore, the S 2p spectrum (Figure 4d

Photocatalytic Performance and Mechanism of Hydrogen Evolution Analysis
The photocatalytic performance of the as-prepared catalysts was evaluated under visible light illumination, whereas the corresponding control experiments were performed without light or the catalyst. For the control experiments, no products were formed, indicating that the light source and photocatalyst were important components for effective photocatalytic H2 evolution. Furthermore, it was found that the parental In2O3, CdO, and CdO@In2O3-25 did not produce H2 gas, as displayed in Figure 5a. The samples doped with sulfur, the S/In2O3 and S/CdO samples, exhibited photocatalytic H2 generation activities with H2 yield rates of 3.6 and 203.4 μmol g −1 h −1 , respectively. As expected, the S/CdO@In2O3-25 catalyst exhibited the best hydrogen production performance with the highest H2 production rate of 4564.5 μmol g −1 h −1 and an ultrahigh H2 yield of 9129.1 μmol·g −1 after a 2 h reaction, as summarized in Figures 5a and S4a. This activity might be attributed to the rich surface oxygen vacancy defects caused by doping S into the hybrid, and forming a type-II heterojunction, which is conducive to the rapid separation/transfer of photo-induced charge carriers. As a result, the H2 evolution rate of S/CdO@In2O3 was higher than that of the S/CdO and S/In2O3 catalysts. Consequently, the as-synthesized hybrid heterojunction was essential for driving the efficient H2 evolution of S/CdO@In2O3. With an increase in the molar ratio of In2O3 to CdO, the H2 evolution rate of the S/CdO@In2O3 hybrid increased (Figure 5a), with a maximum value of 4564. 5

Photocatalytic Performance and Mechanism of Hydrogen Evolution Analysis
The photocatalytic performance of the as-prepared catalysts was evaluated under visible light illumination, whereas the corresponding control experiments were performed without light or the catalyst. For the control experiments, no products were formed, indicating that the light source and photocatalyst were important components for effective photocatalytic H 2 evolution. Furthermore, it was found that the parental In 2 O 3 , CdO, and CdO@In 2 O 3 -25 did not produce H 2 gas, as displayed in Figure 5a. The samples doped with sulfur, the S/In 2 O 3 and S/CdO samples, exhibited photocatalytic H 2 generation activities with H 2 yield rates of 3.6 and 203.4 µmol g −1 h −1 , respectively. As expected, the S/CdO@In 2 O 3 -25 catalyst exhibited the best hydrogen production performance with the highest H 2 production rate of 4564.5 µmol g −1 h −1 and an ultrahigh H 2 yield of 9129.1 µmol·g −1 after a 2 h reaction, as summarized in Figure 5a and Figure S4a. This activity might be attributed to the rich surface oxygen vacancy defects caused by doping S into the hybrid, and forming a type-II heterojunction, which is conducive to the rapid separation/transfer of photo-induced charge carriers. As a result, the H 2 evolution rate of S/CdO@In 2 O 3 was higher than that of the S/CdO and S/In 2 O 3 catalysts. Consequently, the as-synthesized hybrid heterojunction was essential for driving the efficient H 2 evolution of S/CdO@In 2 O 3 . With an increase in the molar ratio of In 2 O 3 to CdO, the H 2 evolution rate of the S/CdO@In 2 O 3 hybrid increased (Figure 5a), with a maximum value of 4564.5 µmol g −1 h −1 H 2 obtained with S/CdO@In 2 O 3 -25. When the In 2 O 3 /CdO molar ratios exceeded 0.25, the H 2 yield rate decreased, which might have been because the stacking of In 2 O 3 accumulation influenced light absorption. This accumulation may lead to the collapse of the material structure. In addition, the H 2 evolution efficiency of S/CdO@In 2 O 3 with various molar ratios of In 2 O 3 /CdO is in good agreement with the PL emission peak strength. After four consecutive recycling reactions, the catalytic activity of S/CdO@In 2 O 3 -25 did not decrease and nearly 85% of the initial activity could be remained (Figure 5b). This strongly implies that the hybrid catalyst exhibited favorable photocatalytic stability, as further evidenced by the FTIR spectra ( Figure S4b). The relationship between the reaction time and generated H 2 amount for the S/CdO@In 2 O 3 -25 was also determined, as shown in Figure S5. When the irradiation is 2 h, the hybrid had the best hydrogen evolution activity with the yield rate of 4564.5 µmol·g −1 ·h −1 .
Nanomaterials 2023, 13, x FOR PEER REVIEW 8 of 13 H2 obtained with S/CdO@In2O3-25. When the In2O3/CdO molar ratios exceeded 0.25, the H2 yield rate decreased, which might have been because the stacking of In2O3 accumulation influenced light absorption. This accumulation may lead to the collapse of the material structure. In addition, the H2 evolution efficiency of S/CdO@In2O3 with various molar ratios of In2O3/CdO is in good agreement with the PL emission peak strength. After four consecutive recycling reactions, the catalytic activity of S/CdO@In2O3-25 did not decrease and nearly 85% of the initial activity could be remained (Figure 5b). This strongly implies that the hybrid catalyst exhibited favorable photocatalytic stability, as further evidenced by the FTIR spectra ( Figure S4b). The relationship between the reaction time and generated H2 amount for the S/CdO@In2O3-25 was also determined, as shown in Figure S5. When the irradiation is 2 h, the hybrid had the best hydrogen evolution activity with the yield rate of 4564.5 μmol·g −1 ·h −1 . UV-Vis diffuse reflectance spectra were also analyzed to determine the origin of the photocatalytic activity of the sulphur-doped hybrid catalysts. As shown in Figure 6a, the absorption values of CdO and In2O3 are approximately 500 and 450 nm, respectively. These values are in good agreement with the values reported in previous studies [54,55]. Compared with the contradistinctive materials, the S/CdO@In2O3-25 composite exhibits a certain red shift. The improved visible light absorption properties may be related to the existence of oxygen vacancies and partial partially valence states. In other words, sulfur doping can effectively optimize the band gap because the hybridization of the O 2p and S 2p orbitals produces additional intermediate electronic states. The band gap was calculated using the Tauc curves, as shown in Figure 6b [56,57]. It can be concluded that the band gap energies (Eg) of S/CdO, S/In2O3 and S/CdO@In2O3-25 were 2.39, 2.54 and 2.22 eV, respectively. Further, photoluminescence (PL) spectroscopy was employed to study the separation of the photoexcited charge carriers. As depicted in Figure 6c, the steady-state PL spectra exhibit a wide peak at approximately 435 nm, resulting from the intrinsic energy band PL and surface oxygen defects. From the as-obtained PL results, it can be clearly seen that the original In2O3 exhibits the strongest PL intensity with a peak at 470 nm, owing to its significant charge recombination. For the CdO@In2O3 hybrid materials, the emission yield of In2O3 decreases significantly, which could be attributed to the strong interaction between In2O3 and CdO that contributed to the more efficient charge separation. By contrast, the S/CdO@In2O3-25 shows the lowest intensity, which strongly indicates that the introduction of S can effectively inhibit charge recombination, consistent with the previous photocatalytic performance. Furthermore, the transient photocurrent responses were determined to analyze the photoexcited charge carrier transfer properties of the materials. As shown in Figure 6d, the S/CdO@In2O3-25 catalyst possesses the highest photocurrent response compared to the other catalyst materials (Table S1), which illustrates that the charge-separation properties in UV-Vis diffuse reflectance spectra were also analyzed to determine the origin of the photocatalytic activity of the sulphur-doped hybrid catalysts. As shown in Figure 6a, the absorption values of CdO and In 2 O 3 are approximately 500 and 450 nm, respectively. These values are in good agreement with the values reported in previous studies [54,55]. Compared with the contradistinctive materials, the S/CdO@In 2 O 3 -25 composite exhibits a certain red shift. The improved visible light absorption properties may be related to the existence of oxygen vacancies and partial partially valence states. In other words, sulfur doping can effectively optimize the band gap because the hybridization of the O 2p and S 2p orbitals produces additional intermediate electronic states. The band gap was calculated using the Tauc curves, as shown in Figure 6b [56,57]. It can be concluded that the band gap energies (Eg) of S/CdO, S/In 2 O 3 and S/CdO@In 2 O 3 -25 were 2.39, 2.54 and 2.22 eV, respectively. Further, photoluminescence (PL) spectroscopy was employed to study the separation of the photoexcited charge carriers. As depicted in Figure 6c, the steady-state PL spectra exhibit a wide peak at approximately 435 nm, resulting from the intrinsic energy band PL and surface oxygen defects. From the as-obtained PL results, it can be clearly seen that the original In 2 O 3 exhibits the strongest PL intensity with a peak at 470 nm, owing to its significant charge recombination. For the CdO@In 2 O 3 hybrid materials, the emission yield of In 2 O 3 decreases significantly, which could be attributed to the strong interaction between In 2 O 3 and CdO that contributed to the more efficient charge separation. By contrast, the S/CdO@In 2 O 3 -25 shows the lowest intensity, which strongly indicates that the introduction of S can effectively inhibit charge recombination, consistent with the previous photocatalytic performance. Furthermore, the transient photocurrent responses were determined to analyze the photoexcited charge carrier transfer properties of the materials. As shown in Figure 6d, the S/CdO@In 2 O 3 -25 catalyst possesses the highest photocurrent response compared to the other catalyst materials (Table S1), which illustrates that the charge-separation properties in the S/CdO@In 2 O 3 -25 hybrid can be largely promoted. In addition, the S/CdO@In 2 O 3 -25 hybrid catalyst has the smallest semicircle compared to the other synthesized catalysts (Figure 6e), indicating that its charge transfer resistance is extremely low, which favors the separation and transfer of photo-induced charge carriers. Nanomaterials 2023, 13, x FOR PEER REVIEW 9 of 13 the S/CdO@In2O3-25 hybrid can be largely promoted. In addition, the S/CdO@In2O3-25 hybrid catalyst has the smallest semicircle compared to the other synthesized catalysts ( Figure  6e), indicating that its charge transfer resistance is extremely low, which favors the separation and transfer of photo-induced charge carriers. The absorption and desorption properties of the H* atom on the photocatalyst play a key role in the H2 evolution reaction. Density functional theory calculations were used to investigate this property in the as-prepared samples. As shown in Figure 6f Table S2. Based on the above results, a type-II charge transfer mechanism is proposed ( Figure 7c)  The absorption and desorption properties of the H* atom on the photocatalyst play a key role in the H 2 evolution reaction. Density functional theory calculations were used to investigate this property in the as-prepared samples. As shown in Figure 6f Table S2. Based on the above results, a type-II charge transfer mechanism is proposed ( Figure 7c). Under simulated light irradiation, S/CdO and S/In 2 O 3 produced excited electron-hole pairs. The generated electrons in S/In 2 O 3 move to the CB of S/CdO, whereas the formed holes migrate to the VB of S/In 2 O 3 , leading to the accumulation of electrons in the CB of S/CdO and h + in the VB of S/In 2 O 3 . In this case, photoreduction of H 2 O to H 2 can be achieved using the appropriate redox potentials of S/CdO. Moreover, the adsorption or desorption behavior (|∆GH*| → 0) of the S/CdO@In 2 O 3 hybrid also facilitates the migration of photogenerated carriers, further promoting the separation of photoinduced electron-hole pairs. irradiation, S/CdO and S/In2O3 produced excited electron-hole pairs. The generated electrons in S/In2O3 move to the CB of S/CdO, whereas the formed holes migrate to the VB of S/In2O3, leading to the accumulation of electrons in the CB of S/CdO and h + in the VB of S/In2O3. In this case, photoreduction of H2O to H2 can be achieved using the appropriate redox potentials of S/CdO. Moreover, the adsorption or desorption behavior (|ΔGH*| → 0) of the S/CdO@In2O3 hybrid also facilitates the migration of photogenerated carriers, further promoting the separation of photoinduced electron-hole pairs.

Conclusions
Overall, a S/CdO@In2O3 hybrid with a ternary heterojunction was successfully synthesized for effective solar-driven H2 evolution. The experimental results illustrated that the S/CdO@In2O3-25 hybrid exhibited a remarkable H2 production rate of 4564.5 μmol g −1 h −1 under artificial sunlight irradiation, and this H2 yield rate was approximately 22.0 and 1261.0 times greater than those of the parental S/CdO and S/In2O3, respectively, surpassing that of many reported photocatalyst materials. Moreover, the as-designed S/CdO@In2O3 hybrid nanofibers and abundant oxygen defects could optimize the electronic structure and activation energies of the catalysts. Furthermore, a double charge-transfer mechanism based on the type-II charge-transfer mechanism was proposed to elucidate the prominent photocatalytic H2 generation activity. This study employed this in situ electrospinning approach to construct 1D S-doped CdO@In2O3 nanofibers for photocatalytic hydrogen evolution. These results of this study can be helpful in the development of efficient solar-to-fuel conversion materials.

Conclusions
Overall, a S/CdO@In 2 O 3 hybrid with a ternary heterojunction was successfully synthesized for effective solar-driven H 2 evolution. The experimental results illustrated that the S/CdO@In 2 O 3 -25 hybrid exhibited a remarkable H 2 production rate of 4564.5 µmol g −1 h −1 under artificial sunlight irradiation, and this H 2 yield rate was approximately 22.0 and 1261.0 times greater than those of the parental S/CdO and S/In 2 O 3 , respectively, surpassing that of many reported photocatalyst materials. Moreover, the as-designed S/CdO@In 2 O 3 hybrid nanofibers and abundant oxygen defects could optimize the electronic structure and activation energies of the catalysts. Furthermore, a double charge-transfer mechanism based on the type-II charge-transfer mechanism was proposed to elucidate the prominent photocatalytic H 2 generation activity. This study employed this in situ electrospinning approach to construct 1D S-doped CdO@In 2 O 3 nanofibers for photocatalytic hydrogen evolution. These results of this study can be helpful in the development of efficient solar-to-fuel conversion materials.