Fabrication of Mn-Doped SrTiO3/Carbon Fiber with Oxygen Vacancy for Enhanced Photocatalytic Hydrogen Evolution

With carbon fiber, it is difficult to load semiconductor photocatalysts and easy to shed off thanks to its smooth surface and few active groups, which has always been a problem in the synthesis of photocatalysts. In the study, SrTiO3 nanoparticles were loaded onto the Tencel fibers using the solvothermal method, and then the Tencel fibers were carbonized at a high temperature under the condition of inert gas to form carbon fibers, thus SrTiO3@CF photocatalytic composite materials with solid core shell structure were prepared. Meanwhile, Mn ions were added into the SrTiO3 precursor reagent in the solvothermal experiment to prepare Mn-doped Mn-SrTiO3@CF photocatalytic composite material. XPS and EPR tests showed that the prepared Mn-SrTiO3@CF photocatalytic composite was rich in oxygen vacancies. The existence of these oxygen vacancies formed oxygen defect states (VOs) below the conduction band, which constituted the capture center of photogenerated electrons and significantly improved the photocatalytic activity. The photocatalytic hydrogen experimental results showed that the photocatalytic hydrogen production capacity of Mn-SrTiO3@CF composite material with 5% Mn-doped was six times that of the SrTiO3@CF material, and the doping of Mn ions not only promoted the red shift of the light absorption boundary and the extension to visible light, but also improved the separation and migration efficiency of photocarriers. In the paper, the preparation method solves the difficulty of loading photocatalysts on CF and provides a new design method for the recycling of catalysts, and we improve the hydrogen production performance of photocatalysts by Mn-doped modification and the introduction of oxygen vacancies, which provides a theoretical method for the practical application of hydrogen energy.


Introduction
As a sustainable green technology, semiconductor photocatalysis can transform solar energy, purify the environment, and produce renewable energy. At present, there are various kinds of semiconductor photocatalysts. Among them, SrTiO 3 semiconductor material has been widely used in fields such as photocatalytic water splitting hydrogen production, virus inactivation, and pollutant treatment thanks to its characteristics of excellent chemical stability, non-toxicity, low cost, good photoelectric property, and environmental friendliness [1]. However, powder SrTiO 3 photocatalytic material is easy to condense into blocks in use, which reduces the photocatalytic property of SrTiO 3 , and causes some problems such as difficulties in the recycling of catalyst, easy consumption, possible secondary pollution, and so on, limiting its practical application. Therefore, some researchers have begun to think about loading photocatalysts on more suitable carriers [2]. Carbon fiber (CF) has

Photocatalytic Evolution
The photocatalytic hydrogen evolution of the samples was tested through a gas chromatograph (GC-7900, Techcomp (China) Co., Ltd., Shanghai, China), which was used to measure the hydrogen production of samples under simulated sunlight irradiation. Before the experiment, 100 mg of catalyst sample was added into the mixture of 50 mL of Na 2 S (0.35 M) solution and 50 mL of Na 2 SO 3 (0.25 M) solution, wherein Na 2 S and Na 2 SO 3 were used as sacrificial agents in the photocatalytic hydrogen production experiment. During the experiment, a PLS-SE300C 300 W xenon lamp was used to simulate the solar light source for testing.

Results and Discussion
The qualitative phase analysis and crystal structure analysis of photocatalysts could be carried out through an x-ray diffraction test. Figure 1 shows the XRD patterns of Mn-SrTiO 3 @CF, SrTiO 3 @CF, carbon fibers, and pure SrTiO 3 . It can be observed from Figure (PDF#35-0734), respectively, which indicates that the loaded SrTiO 3 has a cubic phase perovskite structure [27,28]. Mn-SrTiO 3 @CF have basically the same diffraction peak positions and peak intensities as SrTiO 3 @CF and have no new diffraction peak, from which it can be presumed that the doping of 5% Mn does not significantly change the crystal structure of the composite catalyst [13]. The above test analysis indicated that strontium titanate semiconductor material was successfully loaded on the surface of carbon fibers, but whether Mn was doped into the strontium titanate material needed to be further tested and verified.
doping of 5% Mn does not significantly change the crystal structure of the composite catalyst [13]. The above test analysis indicated that strontium titanate semiconductor material was successfully loaded on the surface of carbon fibers, but whether Mn was doped into the strontium titanate material needed to be further tested and verified. The morphology characteristics of Mn-SrTiO3@CF photocatalytic composite fibers and its preparation process can be further understood through an SEM test and characterization. As shown in Figure 2a,b, Tencel fibers are smooth-faced circular elongated fibers with a diameter of about 10 µm. Figure 2c shows the morphology structure of SrTiO3@Tencel fiber prepared by the solvothermal method; paste-like SrTiO3 particles are coated on the surface of Tencel fibers, and the diameter of SrTiO3@Tencel fiber composite fibers is about 11 µm. It can also be seen from Figure 2c that the SrTiO3 layer on the surface of Tencel fibers has obvious cracks, which may be caused by high pressure irradiation in the process of taking SEM pictures. This is because, under high pressure irradiation, the SrTiO3 material layer would split with the Tencel fiber. In this solvothermal experiment, the Tencel fibers did not dissolve in the high temperature solution because of the use of ethylene glycol and absolute ethyl alcohol as solvents, which effectively prevented cellulose fibers from dissolving. Figure 2d-f shows the morphology structure of SrTiO3@CF composite fibers prepared after the Tencel fibers are carbonized at a high temperature to form carbon fibers, from which it can be observed that SrTiO3 particles are closely coated on the surface of carbon fibers to form a coating layer with a core shell structure. SrTiO3 particles are organically combined with carbon fibers by the high temperature carbonization process to form a firm and close whole. Interestingly, it can be observed from Figure  2f that SrTiO3 nanoparticles have a cubic phase structure, which is consistent with the XRD test results of SrTiO3@CF. In addition, it can be found by comparing Figure 2c and Figure 2d that the diameter of carbon fibers formed by carbonizing Tencel fibers at a high temperature is reduced, and the diameter of the prepared SrTiO3@CF composite fibers is 6-7 µm. Figure 2g-h is an SEM diagram of Mn-SrTiO3@CF composite fibers prepared after doping of 5% Mn, with the surface morphology structure similar to that of SrTiO3@CF composite fibers. Figure 2i shows an SEM diagram of the cross-section structure of Mn-SrTiO3@CF composite fibers, from which it can be observed that the carbon fiber in the core of the composite material has a circular structure, and the thickness of the Mn-SrTiO3 catalyst layer of the shell layer is about 100 nm. The morphology characteristics of Mn-SrTiO 3 @CF photocatalytic composite fibers and its preparation process can be further understood through an SEM test and characterization. As shown in Figure 2a,b, Tencel fibers are smooth-faced circular elongated fibers with a diameter of about 10 µm. Figure 2c shows the morphology structure of SrTiO 3 @Tencel fiber prepared by the solvothermal method; paste-like SrTiO 3 particles are coated on the surface of Tencel fibers, and the diameter of SrTiO 3 @Tencel fiber composite fibers is about 11 µm. It can also be seen from Figure 2c that the SrTiO 3 layer on the surface of Tencel fibers has obvious cracks, which may be caused by high pressure irradiation in the process of taking SEM pictures. This is because, under high pressure irradiation, the SrTiO3 material layer would split with the Tencel fiber. In this solvothermal experiment, the Tencel fibers did not dissolve in the high temperature solution because of the use of ethylene glycol and absolute ethyl alcohol as solvents, which effectively prevented cellulose fibers from dissolving. Figure 2d-f shows the morphology structure of SrTiO 3 @CF composite fibers prepared after the Tencel fibers are carbonized at a high temperature to form carbon fibers, from which it can be observed that SrTiO 3 particles are closely coated on the surface of carbon fibers to form a coating layer with a core shell structure. SrTiO 3 particles are organically combined with carbon fibers by the high temperature carbonization process to form a firm and close whole. Interestingly, it can be observed from Figure 2f that SrTiO 3 nanoparticles have a cubic phase structure, which is consistent with the XRD test results of SrTiO 3 @CF. In addition, it can be found by comparing Figure 2c,d that the diameter of carbon fibers formed by carbonizing Tencel fibers at a high temperature is reduced, and the diameter of the prepared SrTiO 3 @CF composite fibers is 6-7 µm. Figure 2g-h is an SEM diagram of Mn-SrTiO 3 @CF composite fibers prepared after doping of 5% Mn, with the surface morphology structure similar to that of SrTiO 3 @CF composite fibers. Figure 2i shows an SEM diagram of the cross-section structure of Mn-SrTiO 3 @CF composite fibers, from which it can be observed that the carbon fiber in the core of the composite material has a circular structure, and the thickness of the Mn-SrTiO 3 catalyst layer of the shell layer is about 100 nm.
To explore the distribution of various elements in the Mn-SrTiO 3 @CF sample, the SEM mapping test was performed, and the results are shown in Figure 3. It can be observed from Figure 3b that the color of carbon fibers is lighter, which is consistent with the test results in [29]. This is because the SEM mapping test is mainly for the element distribution on the surface of the samples within a certain area, the Mn-SrTiO 3 @CF composite fibers have a core shell structure, and the carbon fibers are in the coated state, so the content of C element detected in the Mn-SrTiO 3 @CF is relatively small. As shown in Figure 3, elements such as Sr, Ti, Mn, and O are evenly distributed on the surface of carbon fibers, and the content of Mn is lower, which corresponds to the actual ratio. To explore the distribution of various elements in the Mn-SrTiO3@CF sample, the SEM mapping test was performed, and the results are shown in Figure 3. It can be observed from Figure 3b that the color of carbon fibers is lighter, which is consistent with the test results in [29]. This is because the SEM mapping test is mainly for the element distribution on the surface of the samples within a certain area, the Mn-SrTiO3@CF composite fibers have a core shell structure, and the carbon fibers are in the coated state, so the content of C element detected in the Mn-SrTiO3@CF is relatively small. As shown in Figure 3, elements such as Sr, Ti, Mn, and O are evenly distributed on the surface of carbon fibers, and the content of Mn is lower, which corresponds to the actual ratio. TEM and HRTEM of the Mn-SrTiO3@CF sample are shown in Figure 4. SrTiO3 plies observed in Figure 4a are nano-fragments shed from the Mn-SrTiO3@CF sample that was cut into pieces, from which it can be seen that most of the SrTiO3 plies are composed of small particles of 10-20 nm combined together, and these small nanoparticles have a larger specific surface area, which is conducive to improving the photocatalytic activity. The corresponding interplanar spacing shown in Figure 4c is about 0.279 nm, which shall belong to the (110) crystal TEM and HRTEM of the Mn-SrTiO 3 @CF sample are shown in Figure 4. SrTiO 3 plies observed in Figure 4a are nano-fragments shed from the Mn-SrTiO 3 @CF sample that was cut into pieces, from which it can be seen that most of the SrTiO 3 plies are composed of small particles of 10-20 nm combined together, and these small nanoparticles have a larger specific surface area, which is conducive to improving the photocatalytic activity. The corresponding interplanar spacing shown in Figure 4c is about 0.279 nm, which shall belong to the (110) crystal face of SrTiO 3 (PDF#35-0734) [30]. The doping of a small amount of Mn may change the spacing of the crystal face, which is not significant. The above experimental results showed that Mn was successfully doped in the SrTiO 3 material [31], and the doping of a small amount of Mn does not significantly change the lattice structure of the SrTiO 3 material. TEM and HRTEM of the Mn-SrTiO3@CF sample are shown in Figure 4. SrTiO3 plies observed in Figure 4a are nano-fragments shed from the Mn-SrTiO3@CF sample that was cut into pieces, from which it can be seen that most of the SrTiO3 plies are composed of small particles of 10-20 nm combined together, and these small nanoparticles have a larger specific surface area, which is conducive to improving the photocatalytic activity. The corresponding interplanar spacing shown in Figure 4c is about 0.279 nm, which shall belong to the (110) crystal face of SrTiO3 (PDF#35-0734) [30]. The doping of a small amount of Mn may change the spacing of the crystal face, which is not significant. The above experimental results showed that Mn was successfully doped in the SrTiO3 material [31], and the doping of a small amount of Mn does not significantly change the lattice structure of the SrTiO3 material.  The composition and valence state of Mn-SrTiO 3 @CF photocatalysts are researched through X-ray photoelectron spectroscopy (XPS). Figure 5 shows the full spectrum of the Mn-SrTiO 3 @CF sample and the high-resolution XPS spectra of C 1s, Sr 3d, Ti 2p, O 1s, and Mn 2p. Figure 5a shows the full spectrum of the Mn-SrTiO 3 @CF sample, indicating that there are mainly Sr, Ti, Mn, O, and C in the prepared sample, without other impurities. In the fitting spectrum of C 1s shown in Figure 5b, the fitting peak at 284.0 eV is a C-C bond, which corresponds to the C-C bond in carbon fibers and is partially similar to the graphite structure, while the fitting peak at 285.2 eV shall belong to a small amount of C-O bonds on the surface of carbon fibers [31]. It is presumed that carbon fibers and SrTiO 3 material are bonded partially through a chemical bond formed by O atoms. It can be observed from Figure 5c that the high resolution XPS spectrum of Ti 2p can be fitted into two characteristic peaks, which are located at the binding energies of 458.0 eV and 463.8 eV, respectively, which, presumably, shall belong to Ti 2p 3/2 and Ti 2p 1/2 , respectively, corresponding to Ti 4+ ions. The difference between the two fitting peaks is 5.8 eV [8]. It can be observed from the spectrum of Sr 3d shown in Figure 5d that peaks at the binding energies of 132.5 eV and 134.3 eV are attributed to Sr 3d 5/2 and Sr 3d 3/2 of Sr 2+ , and the difference between the peaks is 1.8 eV [32]. Interestingly, as shown in Figure 5e, the peak of O 1s is fitted into three peaks that are located at 529.2 eV, 530.9 eV, and 532.4 eV, respectively, wherein the peak at the binding energy of 529.2 eV corresponds to lattice oxygen in SrTiO 3 , while the peak at the binding energy of 530.9 eV may be adsorption oxygen in oxygen vacancy, and the peak at 532.4 eV corresponds to surface adsorption oxygen in the SrTiO 3 catalyst. This test indicated that there may be oxygen vacancies in the Mn-SrTiO 3 @CF catalyst sample [33,34], which needs to be further confirmed by the EPR test. Figure 5f shows the high-resolution XPS diagram of Mn 2p. Because of the low content of doped Mn, the corresponding measured XPS characteristic peak signal is relatively weak. After peak fitting, it is presumed that the characteristic peaks located at the binding energies of 641.6 eV and 652.6 eV correspond to Mn 2p 3/2 and Mn 2p 1/2 , respectively, and the difference between the two peaks is 11 eV. The EPR test was conducted to further confirm the existence of oxygen vacancies in the Mn-SrTiO3@CF sample, as shown in Figure 6. It can be observed from Figure 6 that the Mn-SrTiO3@CF sample has a strong peak at g ≈ 2.003, indicating the existence of oxygen vacancies in the Mn-SrTiO3@CF sample [34,35]. The generation of oxygen vacancies in the Mn-SrTiO3@CF sample may be caused by the fact that, in the carbonization process of bamboo pulp fibers loaded with SrTiO3, a part of C reacts with O in SrTiO3, and after SrTiO3 is treated at high temperature in argon atmosphere, a part of O in the lattices sheds off, thus forming oxygen defects. The existence of these oxygen vacancies will help to expand the light absorption range of the photocatalyst as well as to improve the charge transfer ability of the Mn-SrTiO3@CF material. Combined with the above analysis of results, it indicated that Mn-SrTiO3@CF composite material rich in oxygen vacancies is successfully prepared, and the doping of Mn will further improve the photocatalytic activity of the material. The EPR test was conducted to further confirm the existence of oxygen vacancies in the Mn-SrTiO 3 @CF sample, as shown in Figure 6. It can be observed from Figure 6 that the Mn-SrTiO 3 @CF sample has a strong peak at g ≈ 2.003, indicating the existence of oxygen vacancies in the Mn-SrTiO 3 @CF sample [34,35]. The generation of oxygen vacancies in the Mn-SrTiO 3 @CF sample may be caused by the fact that, in the carbonization process of bamboo pulp fibers loaded with SrTiO 3 , a part of C reacts with O in SrTiO 3 , and after SrTiO 3 is treated at high temperature in argon atmosphere, a part of O in the lattices sheds off, thus forming oxygen defects. The existence of these oxygen vacancies will help to expand the light absorption range of the photocatalyst as well as to improve the charge transfer ability of the Mn-SrTiO 3 @CF material. Combined with the above analysis of results, it indicated that Mn-SrTiO 3 @CF composite material rich in oxygen vacancies is successfully prepared, and the doping of Mn will further improve the photocatalytic activity of the material. The photocatalytic hydrogen production performance of composite materials such as SrTiO3@CF and Mn-SrTiO3@CF was tested under simulated sunlight. The test results are shown in Figure 7. Na2S and Na2SO3 were used as sacrificial agents in this experiment. It can be seen from Figure 7 that carbon fibers have no hydrogen production performance under the action of light, while the hydrogen production performance of SrTiO3@CF photocatalytic composite fibers is about 46.90 μmol/g·h. It is presumed that the existence of oxygen vacancies in the SrTiO3@CF material promotes the extension of its light absorption boundary to visible light and enhances the charge transfer ability, which is conducive to improving the photocatalytic property of the material. Meanwhile, the carbon fibers have functions similar to co-catalysts, promoting the migration of photoelectrons [24]. As shown in Figure 7, the hydrogen production capacity of Mn-SrTiO3@CF composite photocatalytic fibers reaches 285.37 μmol/g·h, which is about six times that of the SrTiO3@CF material, which is obviously attributed to the doping of a small amount of Mn ions; the result is similar to the research in [36]. The doping of Mn ions not only promoted the red shift of the light absorption boundary and the extension to visible light, but also improved the separation and migration efficiency of photocarriers. The above test results showed that the existence of oxygen vacancies, the function of carbon fibers similar to a co-catalyst, and the doping of Mn ions significantly improved the hydrogen production performance of Mn-SrTiO3@CF photocatalytic composite fibers.  The photocatalytic hydrogen production performance of composite materials such as SrTiO 3 @CF and Mn-SrTiO 3 @CF was tested under simulated sunlight. The test results are shown in Figure 7. Na 2 S and Na 2 SO 3 were used as sacrificial agents in this experiment. It can be seen from Figure 7 that carbon fibers have no hydrogen production performance under the action of light, while the hydrogen production performance of SrTiO 3 @CF photocatalytic composite fibers is about 46.90 µmol/g·h. It is presumed that the existence of oxygen vacancies in the SrTiO 3 @CF material promotes the extension of its light absorption boundary to visible light and enhances the charge transfer ability, which is conducive to improving the photocatalytic property of the material. Meanwhile, the carbon fibers have functions similar to co-catalysts, promoting the migration of photoelectrons [24]. As shown in Figure 7, the hydrogen production capacity of Mn-SrTiO 3 @CF composite photocatalytic fibers reaches 285.37 µmol/g·h, which is about six times that of the SrTiO 3 @CF material, which is obviously attributed to the doping of a small amount of Mn ions; the result is similar to the research in [36]. The doping of Mn ions not only promoted the red shift of the light absorption boundary and the extension to visible light, but also improved the separation and migration efficiency of photocarriers. The above test results showed that the existence of oxygen vacancies, the function of carbon fibers similar to a co-catalyst, and the doping of Mn ions significantly improved the hydrogen production performance of Mn-SrTiO 3 @CF photocatalytic composite fibers. The photocatalytic hydrogen production performance of composite materials such as SrTiO3@CF and Mn-SrTiO3@CF was tested under simulated sunlight. The test results are shown in Figure 7. Na2S and Na2SO3 were used as sacrificial agents in this experiment. It can be seen from Figure 7 that carbon fibers have no hydrogen production performance under the action of light, while the hydrogen production performance of SrTiO3@CF photocatalytic composite fibers is about 46.90 μmol/g·h. It is presumed that the existence of oxygen vacancies in the SrTiO3@CF material promotes the extension of its light absorption boundary to visible light and enhances the charge transfer ability, which is conducive to improving the photocatalytic property of the material. Meanwhile, the carbon fibers have functions similar to co-catalysts, promoting the migration of photoelectrons [24]. As shown in Figure 7, the hydrogen production capacity of Mn-SrTiO3@CF composite photocatalytic fibers reaches 285.37 μmol/g·h, which is about six times that of the SrTiO3@CF material, which is obviously attributed to the doping of a small amount of Mn ions; the result is similar to the research in [36]. The doping of Mn ions not only promoted the red shift of the light absorption boundary and the extension to visible light, but also improved the separation and migration efficiency of photocarriers. The above test results showed that the existence of oxygen vacancies, the function of carbon fibers similar to a co-catalyst, and the doping of Mn ions significantly improved the hydrogen production performance of Mn-SrTiO3@CF photocatalytic composite fibers.   The cyclic stability of the Mn-SrTiO 3 @CF composite catalyst was tested, and the results are shown in Figure 8. After four consecutive tests of cyclic photocatalytic water splitting hydrogen production performance, the average photocatalytic hydrogen production performance of the Mn-SrTiO 3 @CF composite catalyst is about 267.69 µmol/g·h. There was only a slight decrease over a period of cyclic experiments, indicating that the the Mn-SrTiO3@CF composite catalyst can maintain relatively stable photocatalytic performance in the water splitting hydrogen production reaction. The cyclic stability of the Mn-SrTiO3@CF composite catalyst was tested, and the results are shown in Figure 8. After four consecutive tests of cyclic photocatalytic water splitting hydrogen production performance, the average photocatalytic hydrogen production performance of the Mn-SrTiO3@CF composite catalyst is about 267.69 μmol/g·h. There was only a slight decrease over a period of cyclic experiments, indicating that the the Mn-SrTiO3@CF composite catalyst can maintain relatively stable photocatalytic performance in the water splitting hydrogen production reaction.  Based on the Kubelka-Munk theory, the band gap (Eg) of semiconductor materials can be calculated according to Equation (1) [38]. As shown in Figure 9b, hν is drawn with (Ahν) 2 , from which it can be obtained that the band gap of single SrTiO3 is about 3.32 eV. Similarly, it can also be calculated that the band gaps of Mn-SrTiO3@CF and SrTiO3@CF composites are about 2.91 eV and 3.08 eV, respectively. The light absorption of catalyst materials such as Mn-SrTiO 3 @CF can be measured by UV-Vis diffuse reflection spectrum test. As shown in Figure 9a, pure SrTiO 3 nanoparticles have an obvious characteristic absorption edge at 375 nm, while the light absorption property of SrTiO 3 @CF is significantly enhanced compared with pure SrTiO 3 , which mainly comes from the strong light absorption property of carbon fibers. Mn-SrTiO 3 @CF photocatalytic composite fibers have stronger light absorption property compared with SrTiO 3 @CF materials, because the doping of Mn reduces the band gap of the composite material and promotes the red shift of the light absorption boundary and the extension to visible light [37]. Mn-SrTiO 3 @CF and SrTiO 3 @CF materials have a significant change in the radian of the bottom of the light absorption edge, which is presumed to be because of the existence of oxygen vacancies in the material, reducing the band gap of the catalyst and further enhancing the light absorption property. The cyclic stability of the Mn-SrTiO3@CF composite catalyst was tested, and the results are shown in Figure 8. After four consecutive tests of cyclic photocatalytic water splitting hydrogen production performance, the average photocatalytic hydrogen production performance of the Mn-SrTiO3@CF composite catalyst is about 267.69 μmol/g·h. There was only a slight decrease over a period of cyclic experiments, indicating that the the Mn-SrTiO3@CF composite catalyst can maintain relatively stable photocatalytic performance in the water splitting hydrogen production reaction. The light absorption of catalyst materials such as Mn-SrTiO3@CF can be measured by UV-Vis diffuse reflection spectrum test. As shown in Figure 9a, pure SrTiO3 nanoparticles have an obvious characteristic absorption edge at 375 nm, while the light absorption property of SrTiO3@CF is significantly enhanced compared with pure SrTiO3, which mainly comes from the strong light absorption property of carbon fibers. Mn-SrTiO3@CF photocatalytic composite fibers have stronger light absorption property compared with SrTiO3@CF materials, because the doping of Mn reduces the band gap of the composite material and promotes the red shift of the light absorption boundary and the extension to visible light [37]. Mn-SrTiO3@CF and SrTiO3@CF materials have a significant change in the radian of the bottom of the light absorption edge, which is presumed to be because of the existence of oxygen vacancies in the material, reducing the band gap of the catalyst and further enhancing the light absorption property. Based on the Kubelka-Munk theory, the band gap (Eg) of semiconductor materials can be calculated according to Equation (1) [38]. As shown in Figure 9b, hν is drawn with (Ahν) 2 , from which it can be obtained that the band gap of single SrTiO3 is about 3.32 eV. Similarly, it can also be calculated that the band gaps of Mn-SrTiO3@CF and SrTiO3@CF composites are about 2.91 eV and 3.08 eV, respectively. Based on the Kubelka-Munk theory, the band gap (E g ) of semiconductor materials can be calculated according to Equation (1) [38]. As shown in Figure 9b, hν is drawn with (Ahν) 2 , from which it can be obtained that the band gap of single SrTiO 3 is about 3.32 eV.
Similarly, it can also be calculated that the band gaps of Mn-SrTiO 3 @CF and SrTiO 3 @CF composites are about 2.91 eV and 3.08 eV, respectively.
where A is absorbance in UV-visible diffuse reflection; hν is photon energy, which is replaced here by the 1024/ wavelength; and C is a constant. The band structure is an important factor affecting the photocatalytic property, and the flat band potential of the semiconductor catalyst can be calculated using the Mott-Schottky equation (Equation (2)) [38,39].
where C is the interface capacitance and V FB is the flat band potential. As shown in Figure 10, the tangent slope of the Mott-Schottky spectral line of SrTiO 3 is positive, indicating that SrTiO 3 is an n-type semiconductor material and the flat band potential of SrTiO 3 is −0.69 eV (calomel electrode, vs. SCE). Based on the fact that the potential of a calomel electrode relative to a standard hydrogen electrode at 25 • C is about 0.24 eV, it can be calculated that the Fermi level corresponding to SrTiO 3 is about −0.45 eV. It is generally believed that the conduction band position of n-type semiconductor is 0.1 eV different from the Fermi level [40], so the conduction band position of SrTiO 3 is −0.55 eV. It can be known from Figure 9b that the band gap of SrTiO 3 is 3.32 eV. It can be calculated that the valence band position of SrTiO 3 is 2.77 eV.  (1) where A is absorbance in UV-visible diffuse reflection; hν is photon energy, which is replaced here by the 1024/ wavelength; and C is a constant.
The band structure is an important factor affecting the photocatalytic property, and the flat band potential of the semiconductor catalyst can be calculated using the Mott-Schottky equation (Equation (2)) [38,39].
where C is the interface capacitance and VFB is the flat band potential. As shown in Figure 10, the tangent slope of the Mott-Schottky spectral line of SrTiO3 is positive, indicating that SrTiO3 is an n-type semiconductor material and the flat band potential of SrTiO3 is −0.69 eV (calomel electrode, vs. SCE). Based on the fact that the potential of a calomel electrode relative to a standard hydrogen electrode at 25 °C is about 0.24 eV, it can be calculated that the Fermi level corresponding to SrTiO3 is about −0.45 eV. It is generally believed that the conduction band position of n-type semiconductor is 0.1 eV different from the Fermi level [40], so the conduction band position of SrTiO3 is −0.55 eV. It can be known from Figure 9b that the band gap of SrTiO3 is 3.32 eV. It can be calculated that the valence band position of SrTiO3 is 2.77 eV. Based on the above research results, we proposed the photocatalytic water splitting hydrogen production mechanism of the Mn-SrTiO3@CF photocatalytic composite material, as shown in Figure 11. Based on the above research results, we proposed the photocatalytic water splitting hydrogen production mechanism of the Mn-SrTiO 3 @CF photocatalytic composite material, as shown in Figure 11.
According to the test results of the UV-Vis diffuse reflection spectrum (Figure 9), it can be calculated that the composite band gaps of the pure SrTiO 3 , Mn-SrTiO 3 @CF, and SrTiO 3 @CF composite materials are about 3.32 eV, 2.91 eV, and 3.08 eV, respectively. The results indicated that the doping of Mn and the existence of oxygen vacancies reduced the band gap of the corresponding material; expanded the light absorption range, which extended to the visible light; and improved the photocatalytic activity. The Mn-SrTiO 3 @CF composite material has a large number of oxygen vacancies, which will create a new donor level below the conduction band, constituting an oxygen vacancy state (VOs) and becoming the capture center of photoelectrons. As shown in Figure 11, under the action of light, the Mn-SrTiO 3 composite material produces electrons and holes, and photoelectrons migrate to the conduction band and the oxygen vacancy state (VOs) [41]. Some of the electrons located on the conduction band will migrate to the surface of carbon fibers, and some will migrate to the oxygen vacancy state (VOs), which promotes the separation and migration of photoelectrons and holes, thus improving the photocatalytic property. The electrons on the conduction band and the oxygen vacancy state (VOs) will combine with H + ions in water to produce hydrogen, while the holes on the valence band will combine with sacrificial agents (Na 2 S and Na 2 SO 3 ) in aqueous solution to promote the separation and generation of photoelectron-hole pairs. According to the test results of the UV-Vis diffuse reflection spectrum (Figure 9), it can be calculated that the composite band gaps of the pure SrTiO3, Mn-SrTiO3@CF, and SrTiO3@CF composite materials are about 3.32 eV, 2.91 eV, and 3.08 eV, respectively. The results indicated that the doping of Mn and the existence of oxygen vacancies reduced the band gap of the corresponding material; expanded the light absorption range, which extended to the visible light; and improved the photocatalytic activity. The Mn-SrTiO3@CF composite material has a large number of oxygen vacancies, which will create a new donor level below the conduction band, constituting an oxygen vacancy state (VOs) and becoming the capture center of photoelectrons. As shown in Figure 11, under the action of light, the Mn-SrTiO3 composite material produces electrons and holes, and photoelectrons migrate to the conduction band and the oxygen vacancy state (VOs) [41]. Some of the electrons located on the conduction band will migrate to the surface of carbon fibers, and some will migrate to the oxygen vacancy state (VOs), which promotes the separation and migration of photoelectrons and holes, thus improving the photocatalytic property. The electrons on the conduction band and the oxygen vacancy state (VOs) will combine with H + ions in water to produce hydrogen, while the holes on the valence band will combine with sacrificial agents (Na2S and Na2SO3) in aqueous solution to promote the separation and generation of photoelectron-hole pairs.

Conclusions
In this work, Tencel fibers were taken as the substrate, and SrTiO3@CF and Mn-SrTiO3@CF with a firm structure were successfully obtained through the process route of first loading the semiconductor material on the carrier and then carbonizing the tencel fibers. This solved the problem that the semiconductor materials were difficult to directly load on the surfaces of carbon fibers or easy to shed off because of the smooth surface and few active groups of carbon fibers. The Mn-SrTiO3@CF composite photocatalytic fibers Figure 11. Hydrogen production mechanism of Mn-SrTiO 3 @CF photocatalytic composite material.

Conclusions
In this work, Tencel fibers were taken as the substrate, and SrTiO 3 @CF and Mn-SrTiO 3 @CF with a firm structure were successfully obtained through the process route of first loading the semiconductor material on the carrier and then carbonizing the tencel fibers. This solved the problem that the semiconductor materials were difficult to directly load on the surfaces of carbon fibers or easy to shed off because of the smooth surface and few active groups of carbon fibers. The Mn-SrTiO 3 @CF composite photocatalytic fibers exhibited a higher activity for hydrogen evolution compared with the SrTiO 3 @CF material. Particularly, the photocatalytic hydrogen production of the Mn-SrTiO 3 @CF composite catalyst is about 267.69 µmol/g·h with 5% Mn-doped, which is six times that of the SrTiO 3 @CF material. After modifying SrTiO 3 @CF with Mn, the light absorption boundary could be extended to the visible light direction, and the separation and migration efficiency of photocarriers could be improved. In addition, SrTiO 3 @CF and Mn-SrTiO 3 @CF photocatalytic materials rich in oxygen vacancies were successfully prepared through the high temperature carbonization process. The existence of oxygen vacancies would generate a new donor level below the conduction band, constituting an oxygen vacancy state (VOs) and becoming the capture center of photoelectrons, thus significantly improving the photocatalytic activity. The synergistic effect from Mn doping, oxygen vacancies, the sacrificial agent, and carbon fibers can efficiently absorb photons, transfer photoinduced electrons, restrain carrier recombination, and improve the efficiency of the catalyst hydrogen production.