Photocatalytic Properties of a Novel Keratin char-TiO 2 Composite Films Made through the Calcination of Wool Keratin Coatings Containing TiO 2 Precursors

: In this study, the photocatalytic properties of novel keratin char-TiO 2 composite ﬁlms, made through the calcination of wool keratin coatings containing TiO 2 precursors at 400 ◦ C, were investigated for the photodegradation of organic contaminants under visible light irradiation. Its structural characteristics and photocatalytic performance were systematically examined. It was shown that a self-cleaning hydrophobic keratin char-TiO 2 composite ﬁlm containing meso- and micro-pores was formed after the keratin—TiO 2 precursors coating was calcined. In comparison with calcinated TiO 2 ﬁlms, the keratin char-TiO 2 composite ﬁlms doped with the elements of C, N, and S from keratins resulted in decreased crystallinity and a larger water contact angle. The bandgap of the char-TiO 2 composite ﬁlms increased slightly from 3.26 to 3.32 eV, and its separation of photogenerated charge carriers was inhibited to a certain degree. However, it exhibited higher photodegradation efﬁciency to methyl blue (MB) efﬂuents than the pure calcinated TiO 2 ﬁlms. This was mainly because of its special porous structure, large water contact angle, and high adsorption energy towards organic pollutants, conﬁrmed by the density functional theory calculations. The main active species were 1 O 2 radicals in the MB photodegradation process. 2 2 to 2 The annealing showed that the of the TiO 2 ﬁlms and keratin char-TiO 2 composite ﬁlms were in the sub-micrometer range with good adherence to the The magniﬁed FESEM images of their cross-sections and surfaces demonstrated that many meso- and micro-porous holes existed in both the internal and surface of all the four ﬁlms, which was in agreement with TiO 2 ﬁlm generated from the sol-gel process After calcination, more pores were produced in the keratin char-TiO 2 composite ﬁlm. ﬁlms,


Introduction
The semiconductor TiO 2 can be used in a wide range of applications such as paints, cosmetics, photoelectric sensors, energy conversion and storage, air and water purifications, and removal of organic pollutants [1]. With respect to photocatalysis, it is considered one of the most excellent photocatalytic materials due to its low cost, less toxicity, good physical and chemical stabilities, and high photocatalytic activities [2]. In general, TiO 2 has three phase structures, namely rutile, anatase and brookite, and anatase and rutile TiO 2 can be applied in the field of photocatalysis [3]. The theoretical band gaps are 3.2 eV for anataseTiO 2 , and 3.0 eV for rutileTiO 2 , and their corresponding absorption edges are separately located at 388 nm and 414 nm [4]. In comparison with rutile TiO 2 , anatase TiO 2 exhibits stronger photocatalytic activity due to its high local state density, abundant surface hydroxyl radicals, and long lifetime photogenerated charge carriers [5]. However, both can only be excited by ultraviolet rays [6], and their photocatalytic activities are impaired to some degree due to the rapid recombination of photogenerated electron-hole pairs [7]. However, the decreased particle size of TiO 2 particles can lead to the agglomeration phenomenon, which will hinder the incidence of light into the active center [8]. In addition, the separation, recovery, and reuse of nano-scaled TiO 2 photocatalysts pose many environmental, health, and safety problems in practical application [9]. catalysts could still be doped with the elements of C, N, S from keratin and might still have higher efficiency while resistant to harsh alkaline effluents conditions. The keratin hydrogel-TiO 2 composites have been developed for the high-efficiency adsorption and photodegradation of trimethoprim under UV and visible light irradiation [33]. TiO 2 particles are dispersed in polylactic acid (PLA) solution in the presence of feather keratin and clay to prepare the keratin/PLA/TiO 2 /clay nanofibers. The as-prepared nanofibers not only adsorb and remove the MB dye from the aqueous solution but also degrade the MB dye under UV irradiation [34]. Currently, neither keratin char-TiO 2 nanoparticle composite films have been investigated, nor have the porous films composed of TiO2 nanoparticles doped with the elements C, N, and S from keratins char been reported.
In the present work, a novel type of keratin char-TiO 2 composite films doped with C, N, and S elements from wool keratins was prototyped through the calcination of keratin coatings containing TiO 2 precursors. The photocatalytic structures and properties of the keratin char-TiO 2 composite films were investigated, and its photodegradation efficiency to MB under visible light irradiation was reported. Wool keratin was extracted from wool fibers firstly using reducing agents. A spin-coating technique was then alternatively applied to deposit wool keratin and TiO 2 precursors films on quartz glass substrates, followed by annealing at 400 • C. The morphology, surface roughness, phase structure, microstructure, chemical bonding state, hydrophilic properties, energy band structures, and separation of photoinduced charge carriers of the as-prepared keratin char-TiO 2 composite films were analyzed. The photocatalytic mechanism of its photodegradation to MB dyes was elucidated. The role of wool keratin's C, N, and S elements on the photocatalytic properties of keratin char-TiO 2 composite films were verified by density functional theory (DFT) calculations.

Molecular Weight of Wool Keratins
The electrophoretic mobilities of wool keratins and marker proteins were compared to estimate the molecular mass of as-prepared wool keratins [35]. The electrophoretic separation patterns of standard and wool keratins are depicted in Figure 1. The molecular weight from 14 to 97 kDa was applied to line a of Figure 1 for calibration. The electrophoretic patterns of wool keratin (lines 1 and 2) indicated that the broad dull bands with a low molecular weight of 31-14.4 kDa were ascribed to the high sulfur proteins of wool keratin. In contrast, the high molecular weight bands (66.2-43 kDa) correlated with the wool cortex's low-sulfur intermediate filaments proteins having a typical α-helix secondary structure [28].
dyes under visible light irradiation [31]. Porous wool keratin film can be applied for filtration, air purification, and tissue engineering [32]. However, TiO2 nanoparticles doped with wool keratins might be prone to alkaline effluents due to the nature of keratin. It is envisaged that keratin char-TiO2 nanoparticle composite catalysts could still be doped with the elements of C, N, S from keratin and might still have higher efficiency while resistant to harsh alkaline effluents conditions. The keratin hydrogel-TiO2 composites have been developed for the high-efficiency adsorption and photodegradation of trimethoprim under UV and visible light irradiation [33]. TiO2 particles are dispersed in polylactic acid (PLA) solution in the presence of feather keratin and clay to prepare the keratin/PLA/TiO2/clay nanofibers. The as-prepared nanofibers not only adsorb and remove the MB dye from the aqueous solution but also degrade the MB dye under UV irradiation [34]. Currently, neither keratin char-TiO2 nanoparticle composite films have been investigated, nor have the porous films composed of TiO2 nanoparticles doped with the elements C, N, and S from keratins char been reported.
In the present work, a novel type of keratin char-TiO2 composite films doped with C, N, and S elements from wool keratins was prototyped through the calcination of keratin coatings containing TiO2 precursors. The photocatalytic structures and properties of the keratin char-TiO2 composite films were investigated, and its photodegradation efficiency to MB under visible light irradiation was reported. Wool keratin was extracted from wool fibers firstly using reducing agents. A spin-coating technique was then alternatively applied to deposit wool keratin and TiO2 precursors films on quartz glass substrates, followed by annealing at 400 °C. The morphology, surface roughness, phase structure, microstructure, chemical bonding state, hydrophilic properties, energy band structures, and separation of photoinduced charge carriers of the as-prepared keratin char-TiO2 composite films were analyzed. The photocatalytic mechanism of its photodegradation to MB dyes was elucidated. The role of wool keratin's C, N, and S elements on the photocatalytic properties of keratin char-TiO2 composite films were verified by density functional theory (DFT) calculations.

Molecular Weight of Wool Keratins
The electrophoretic mobilities of wool keratins and marker proteins were compared to estimate the molecular mass of as-prepared wool keratins [35]. The electrophoretic separation patterns of standard and wool keratins are depicted in Figure 1. The molecular weight from 14 to 97 kDa was applied to line a of Figure 1 for calibration. The electrophoretic patterns of wool keratin (lines 1 and 2) indicated that the broad dull bands with a low molecular weight of 31-14.4 kDa were ascribed to the high sulfur proteins of wool keratin. In contrast, the high molecular weight bands (66.2-43 kDa) correlated with the wool cortex's low-sulfur intermediate filaments proteins having a typical α-helix secondary structure [28].

Morphology and Surface Roughness of As-Prepared Keratin char-TiO 2 Films
The FESEM and laser microscopy images of the TiO 2 films and keratin char-TiO 2 composite films before and after calcination are displayed in Figure 2. It was noticed that a compact coating was homogeneously deposited on the glass substrate for all four films. After calcination, the surface roughness increased from 1.0 to 3.2 nm for the TiO 2 films and from 3.2 to 9.7 nm for the keratin char-TiO 2 composite films with spot-like features. This was because of the crystallization growth of TiO 2 grains with phase changes from amorphous to anatase at 400 • C [36]. The surface roughness of the keratinchar-TiO 2 composite films was larger than that of the TiO 2 ones, suggesting the macromolecular chains of wool keratin could lead to the agglomeration of TiO 2 . The annealing treatment accelerated the crystal growth and agglomeration of TiO 2 nanoparticles with improved crystallinity [20], contrary to the previous reports [37]. The cross-section FESEM images showed that the thicknesses of the TiO 2 films and keratin char-TiO 2 composite films were in the sub-micrometer range with good adherence to the substrate. The magnified FESEM images of their cross-sections and surfaces demonstrated that many meso-and microporous holes existed in both the internal and surface of all the four films, which was in agreement with TiO 2 film generated from the sol-gel process [38]. After calcination, more pores were produced in the keratin char-TiO 2 composite film.

Morphology and Surface Roughness of As-Prepared Keratin char-TiO2 Films
The FESEM and laser microscopy images of the TiO2 films and keratin char-TiO2 composite films before and after calcination are displayed in Figure 2. It was noticed that a compact coating was homogeneously deposited on the glass substrate for all four films. After calcination, the surface roughness increased from 1.0 to 3.2 nm for the TiO2 films and from 3.2 to 9.7 nm for the keratin char-TiO2 composite films with spot-like features. This was because of the crystallization growth of TiO2 grains with phase changes from amorphous to anatase at 400 °C [36]. The surface roughness of the keratinchar-TiO2 composite films was larger than that of the TiO2 ones, suggesting the macromolecular chains of wool keratin could lead to the agglomeration of TiO2. The annealing treatment accelerated the crystal growth and agglomeration of TiO2 nanoparticles with improved crystallinity [20], contrary to the previous reports [37]. The cross-section FESEM images showed that the thicknesses of the TiO2 films and keratin char-TiO2 composite films were in the sub-micrometer range with good adherence to the substrate. The magnified FESEM images of their cross-sections and surfaces demonstrated that many meso-and micro-porous holes existed in both the internal and surface of all the four films, which was in agreement with TiO2 film generated from the sol-gel process [38]. After calcination, more pores were produced in the keratin char-TiO2 composite film. Figure 2. Images of FESEM and laser microscopy for the TiO2 films (a1-a4) before and (b1-b4) after calcination, and for the keratin char-TiO2 composite films (c1-c4) before and (d1-d4) after calcination (The numbers of 1, 2, 3 and 4 represent the cross-section, magnified cross-section and surface, and three-dimensional micrograph of the films, respectively).

Crystal Structures
The XRD patterns of the TiO 2 films and keratin char-TiO 2 composite films before and after calcination, as well as TiO 2 particles, are manifested in Figure 3. The broad diffraction peak at around 2θ = 22 • in the XRD patterns for all the films belonged to the characteristic peak of quartz glass [39]. Prior to annealing, no diffraction peak could be found in the patterns of the TiO 2 and keratin char-TiO 2 films, implying TiO 2 was an amorphous structure. After calcination, a relatively weak diffraction peak at 25.1 • was detected in the patterns of the calcined films because of the low loading of TiO 2 . This was attributed to the (101) crystal plane of anatase phase TiO 2 (JCPDS No. 21-1272). The peak intensity of the calcined TiO 2 film was stronger than that of the calcined keratin char-TiO 2 composite film along the preferred (101) plane growth orientation. Thus, the incorporation of wool keratin into TiO 2 affected the crystallization of TiO 2 to a certain degree. For comparison, the XRD pattern of pure TiO 2 nanoparticles indicated that the characteristic diffraction peaks at 25 [40]. According to the Scherrer formula [41], the mean crystal size of pure TiO 2 nanoparticles was calculated to be 9.6 nm by measuring the diffraction peak widths of (101), (004), and (200) crystal planes.

Crystal Structures
The XRD patterns of the TiO2 films and keratin char-TiO2 composite films before and after calcination, as well as TiO2 particles, are manifested in Figure 3. The broad diffraction peak at around 2θ = 22° in the XRD patterns for all the films belonged to the characteristic peak of quartz glass [39]. Prior to annealing, no diffraction peak could be found in the patterns of the TiO2 and keratin char-TiO2 films, implying TiO2 was an amorphous structure. After calcination, a relatively weak diffraction peak at 25.1° was detected in the patterns of the calcined films because of the low loading of TiO2. This was attributed to the (101) crystal plane of anatase phase TiO2 (JCPDS No.21-1272). The peak intensity of the calcined TiO2 film was stronger than that of the calcined keratin char-TiO2 composite film along the preferred (101) plane growth orientation. Thus, the incorporation of wool keratin into TiO2 affected the crystallization of TiO2 to a certain degree. For comparison, the XRD pattern of pure TiO2 nanoparticles indicated that the characteristic diffraction peaks at 25.1°, 37.8°, 47.8°, 54.2°, 55.4°, 62.4°, and 75.3° were well matched with the (101), (004), (200), (105), (211), (204), (116), (220), and (215) crystal planes of tetragonal anatase TiO2 [40]. According to the Scherrer formula [41], the mean crystal size of pure TiO2 nanoparticles was calculated to be 9.6 nm by measuring the diffraction peak widths of (101), (004), and (200) crystal planes.

Microstructures of the Calcined Films
The TEM images and selected area electron diffraction (SAED) patterns of the calcined TiO2 film and calcined keratin char-TiO2 composite film are presented in Figure 4. It was seen that after calcination, the crystalline TiO2 particles were formed, and the particle sizes for both films were less than 10 nm with a uniform distribution, which were well consistent with the calculated values from XRD. The crystal lattice spacing in highresolution TEM images were estimated to be 0.32 nm for both the calcined TiO2 film and calcined keratin char-TiO2 composite film, which were close to 0.35 nm of anatase TiO2 [42]. Moreover, the crystalline TiO2 nanoparticles were surrounded by disordered carbon layers. The carbon-type structure was in-situ formed in the TiO2 films calcined at 400 °C because of the oxidation of hydrocarbon compounds in air. However, the lattice spacing of carbon was not found primarily because the calcination temperature was not high enough to produce the crystal structure of carbon [43]. As shown in SAED patterns, the intensities of the diffraction rings of calcined TiO2 nanoparticles were stronger than those of the calcined keratin char-TiO2 ones. This implied that the crystallinity of the calcined TiO2 nanoparticles in keratin char-TiO2 composite film was not as great as pure TiO2 particles. Namely, the smaller crystallinity was probably caused by a mixture of TiO2 and amorphous char/carbon structures. As a result, the in-situ hybridization of C with TiO2 for both the calcined films might improve the separation of photoinduced electron and hole pairs via the electronic interaction between C and TiO2 crystals in the films. Besides,

Microstructures of the Calcined Films
The TEM images and selected area electron diffraction (SAED) patterns of the calcined TiO 2 film and calcined keratin char-TiO 2 composite film are presented in Figure 4. It was seen that after calcination, the crystalline TiO 2 particles were formed, and the particle sizes for both films were less than 10 nm with a uniform distribution, which were well consistent with the calculated values from XRD. The crystal lattice spacing in high-resolution TEM images were estimated to be 0.32 nm for both the calcined TiO 2 film and calcined keratin char-TiO 2 composite film, which were close to 0.35 nm of anatase TiO 2 [42]. Moreover, the crystalline TiO 2 nanoparticles were surrounded by disordered carbon layers. The carbon-type structure was in-situ formed in the TiO 2 films calcined at 400 • C because of the oxidation of hydrocarbon compounds in air. However, the lattice spacing of carbon was not found primarily because the calcination temperature was not high enough to produce the crystal structure of carbon [43]. As shown in SAED patterns, the intensities of the diffraction rings of calcined TiO 2 nanoparticles were stronger than those of the calcined keratin char-TiO 2 ones. This implied that the crystallinity of the calcined TiO 2 nanoparticles in keratin char-TiO 2 composite film was not as great as pure TiO 2 particles. Namely, the smaller crystallinity was probably caused by a mixture of TiO 2 and amorphous char/carbon structures. As a result, the in-situ hybridization of C with TiO 2 for both the calcined films might improve the separation of photoinduced electron and hole pairs via the electronic interaction between C and TiO 2 crystals in the films. Besides, the N and S

Surface Contact Angles
The optical images of the water contact angles for all the films are represented in Figure 5. The water contact angle of the blank quartz glass was 58.0°. The hydrophilicity of the calcined TiO2 film became better, and its contact angle decreased to 36.2° from 90.2° after calcination at 400 °C. The hydrophilic surface might be due to the removal of organic components, the structure change from amorphous to anatase phase, and the increased hydroxyl groups [44]. When wool keratin was introduced into TiO2, the contact angle of the keratin char-TiO2 composite film increased to 96.6° because of the hydrophobic features of wool keratin and the special topography of film [45]. However, the UV irradiation could improve the keratin char-TiO2 composite film's hydrophilicity to some extent with a water contact angle of 72.3°. The water contact angle of the calcined keratin char-TiO2 composite film decreased to 61.7°, implying the limited enhanced wettability. Therefore, the hydrophilic behavior of the films played a positive role in the photodegradation of the MB dye solution [44].

Surface Contact Angles
The optical images of the water contact angles for all the films are represented in Figure 5. The water contact angle of the blank quartz glass was 58.0 • . The hydrophilicity of the calcined TiO 2 film became better, and its contact angle decreased to 36.2 • from 90.2 • after calcination at 400 • C. The hydrophilic surface might be due to the removal of organic components, the structure change from amorphous to anatase phase, and the increased hydroxyl groups [44]. When wool keratin was introduced into TiO 2 , the contact angle of the keratin char-TiO 2 composite film increased to 96.6 • because of the hydrophobic features of wool keratin and the special topography of film [45]. However, the UV irradiation could improve the keratin char-TiO 2 composite film's hydrophilicity to some extent with a water contact angle of 72.3 • . The water contact angle of the calcined keratin char-TiO 2 composite film decreased to 61.7 • , implying the limited enhanced wettability. Therefore, the hydrophilic behavior of the films played a positive role in the photodegradation of the MB dye solution [44].

Surface Contact Angles
The optical images of the water contact angles for all the films are represented in Figure 5. The water contact angle of the blank quartz glass was 58.0°. The hydrophilicity of the calcined TiO2 film became better, and its contact angle decreased to 36.2° from 90.2° after calcination at 400 °C. The hydrophilic surface might be due to the removal of organic components, the structure change from amorphous to anatase phase, and the increased hydroxyl groups [44]. When wool keratin was introduced into TiO2, the contact angle of the keratin char-TiO2 composite film increased to 96.6° because of the hydrophobic features of wool keratin and the special topography of film [45]. However, the UV irradiation could improve the keratin char-TiO2 composite film's hydrophilicity to some extent with a water contact angle of 72.3°. The water contact angle of the calcined keratin char-TiO2 composite film decreased to 61.7°, implying the limited enhanced wettability. Therefore, the hydrophilic behavior of the films played a positive role in the photodegradation of the MB dye solution [44].

Chemical Bonds and Element Analyses
The XPS spectra of the TiO 2 films and keratin char-TiO 2 composite films before and after calcination are revealed in Figure 6. The XPS survey spectra and chemical elements confirmed the doping of N and S elements of keratins withTiO 2 . The relatively high contents of the C element in both the uncalcined films (TiO 2 40.60%, keratin-TiO 2 53.22%) suggested that the organic carbon layers acting as the photosensitizer existed in the boundaries of TiO 2 grains, which would result in the interstitial carbon doping into TiO 2 [46]. The intensities of Ti and O elements in the TiO 2 films and keratin char-TiO 2 composite films were distinctly enhanced after calcination owing to the loss of organic substances at high temperatures. For the stoichiometric formation of TiO 2 ,after calcination, the atomic concentration ratios of O to Ti were reduced from 2.59 to 2.30 for the TiO 2 films and from 3.78 to 2.27 for the keratin char-TiO 2 composite films. The corresponding atomic concentrations of the C element decreased to 19.56% for the TiO 2 films and 22.80% for the keratin char-TiO 2 composite films.
The high-resolution XPS spectra indicated that C=C (carbide)andC-Ti 4+ /Ti 3+ , C-C/C-H, and C-O groups existed in the C 1s spectra of all the films. The number of active sites might be increased because of the introduction of C element into the lattice of TiO 2, which was caused by the oxidation and decomposition of organic matter at 400 • C [47]. It meant the Ti-C or Ti-O-C bonds between the Ti of the TiO 2 and carbon content were formed at this annealing temperature [48]. Hence, C doping could enhance the photocatalytic properties of TiO 2 film. The groups of O-Ti 3+ (lattice oxygen), O-Ti 4+ (TiO 2 ), and O-C (caused by organic components from isopropyl alcohol, acetic acid, and TTIP) presented in the O 1s spectra of all the films. Additionally, Ti 3+ 2p1/2 -O/C groups were fitted in the Ti 2p spectra of all the films [17]. Ti 3+ acting as an active site for oxygen desorption was assigned to the partial substitution of O atoms by C atoms in TiO 2 , promoting the separation of photogenerated electron and hole pairs [13].
In the cases of the keratin char-TiO 2 composite films, besides the above-mentioned fitted subpeaks in the C 1s , O 1s, and Ti 2p spectra, C-N and C=O groups were fitted in the C 1s spectra, and the corresponding O=C group was deconvoluted in the O 1s spectra. Likely, the N 1s XPS spectrum could be deconvoluted into N-C and N-Ti 4+ /Ti 3+ (substitutional and interstitial states of N atom in TiO 2 lattice) [46], and the S 2p spectrum was fitted into S-S, S-H, and S-Ti 4+ /Ti 3+ [49]. After calcination, the S-H group disappeared, but the other two S-S and S-Ti 4+ /Ti 3+ were identified. Furthermore, the fitted groups of Ti 3+ 2p3/2 -O/C/N/S, Ti 3+ 2p1/2 -O/C/N/S, Ti 4+ 2p3/2 -O/C/N/S, and Ti 4+ 2p1/2 -O/C/N/S were well consistent with the high-resolution XPS spectra of C 1s , O 1s , N 1s , S 2p, and Ti 2p . Consequently, the multi-element doping of C, N, and S elements in carbon layers was realized in the keratin char-TiO 2 composite films, and the calcination treatment was helpful for the crystallization growth of C, N, and S co-doped TiO 2 grains while the structure was changed from amorphous to anatase phase [50]. In comparison with S-doped TiO 2 , the Cdoped TiO 2 possessed a higher photocatalytic activity in the dye degradation process [13].

Chemical Bonds and Element Analyses
The XPS spectra of the TiO2 films and keratin char-TiO2 composite films before and after calcination are revealed in Figure 6. The XPS survey spectra and chemical elements confirmed the doping of N and S elements of keratins withTiO2. The relatively high contents of the C element in both the uncalcined films (TiO2 40.60%, keratin-TiO2 53.22%) suggested that the organic carbon layers acting as the photosensitizer existed in the boundaries of TiO2 grains, which would result in the interstitial carbon doping into TiO2 [46]. The intensities of Ti and O elements in the TiO2 films and keratin char-TiO2 composite films were distinctly enhanced after calcination owing to the loss of organic substances at high temperatures. For the stoichiometric formation of TiO2,after calcination,the atomic concentration ratios of O to Ti were reduced from 2.59 to 2.30 for the TiO2 films and from 3.78 to 2.27 for the keratin char-TiO2 composite films. The corresponding atomic concentrations of the C element decreased to 19.56% for the TiO2 films and 22.80% for the keratin char-TiO2 composite films.
The high-resolution XPS spectra indicated that C=C (carbide)andC-Ti 4+ /Ti 3+ , C-C/C-H, and C-O groups existed in the C1s spectra of all the films. The number of active sites might be increased because of the introduction of C element into the lattice of TiO2, which was caused by the oxidation and decomposition of organic matter at 400 °C [47]. It meant the Ti-C or Ti-O-C bonds between the Ti of the TiO2 and carbon content were formed at this annealing temperature [48]. Hence, C doping could enhance the photocatalytic properties of TiO2 film. The groups of O-Ti 3+ (lattice oxygen), O-Ti 4+ (TiO2), and O-C (caused by organic components from isopropyl alcohol, acetic acid, and TTIP) presented in the O1s spectra of all the films. Additionally, Ti 3+ 2p3/2-O/C, Ti 3+ 2p1/2-O/C, Ti 4+ 2p3/2-O/C, and Ti 4+ 2p1/2-O/C groups were fitted in the Ti2p spectra of all the films [17]. Ti 3+ acting as an active site for oxygen desorption was assigned to the partial substitution of O atoms by C atoms in TiO2, promoting the separation of photogenerated electron and hole pairs [13].
In the cases of the keratin char-TiO2 composite films, besides the above-mentioned fitted subpeaks in the C1s, O1s, and Ti2p spectra, C-N and C=O groups were fitted in the C1s spectra, and the corresponding O=C group was deconvoluted in the O1s spectra. Likely, the N1s XPS spectrum could be deconvoluted into N-C and N-Ti 4+ /Ti 3+ (substitutional and interstitial states of N atom in TiO2 lattice) [46], and the S2p spectrum was fitted into S-S, S-H, and S-Ti 4+ /Ti 3+ [49]. After calcination, the S-H group disappeared, but the other two S-S and S-Ti 4+ /Ti 3+ were identified. Furthermore, the fitted groups of Ti 3+ 2p3/2-O/C/N/S, Ti 3+ 2p1/2-O/C/N/S, Ti 4+ 2p3/2-O/C/N/S, and Ti 4+ 2p1/2-O/C/N/S were well consistent with the high-resolution XPS spectra of C1s, O1s, N1s, S2p, and Ti2p. Consequently, the multi-element doping of C, N, and S elements in carbon layers was realized in the keratin char-TiO2 composite films, and the calcination treatment was helpful for the crystallization growth of C, N, and S co-doped TiO2 grains while the structure was changed from amorphous to anatase phase [50]. In comparison with S-doped TiO2, the C-doped TiO2 possessed a higher photocatalytic activity in the dye degradation process [13].

Energy Band Gap Structures
The transmission spectra inserted with the plots of (αhν) 1/2 versus hν in the 200-800 nm range, the UPS spectra, and the corresponding band structures of the TiO 2 films and keratin char-TiO 2 composite films are illustrated in Figure 7. The transmittances for all the films sharply decreased at the light absorption edge around 300 nm. The excitation of electrons induced the strong absorption in UV region from its valence band to the conduction band of TiO 2 . After calcination, the absorption edges of the calcined TiO 2 film and keratin char-TiO 2 composite film were shifted towards longer wavelengths. This was attributed to the structure change from amorphous to anatase phase for both films, the C doping for the TiO 2 film and the C, N, and S co-doping for the keratin char-TiO 2 composite film [51]. All the films exhibited good transmittance in visible light wavebands.
were separately estimated to be 15.49 and 17.52 eV by extrapolating the straight portion of the UPS curve at high binding energy region to zero. It was demonstrated that the work function of semiconductors could be calculated by subtracting the cutoff energy from the photon energy of the He I light source (21.2 eV) [53]. Thus, the work functions of the-calcinedTiO2 film and calcined keratin char-TiO2 composite film were 5.71 and 3.68 eV with respect to the vacuum level. The Fermi energies of thecalcinedTiO2 film and calcined keratin char-TiO2 composite film were −5.71 and −3.68 eV under vacuum conditions, respectively. Moreover, the valence band energy was estimated by extrapolating the straight line of the UPS curve at the low binding energy region to zero, which was lower than its Fermi energy. The valence band energies versus vacuum level were respectively calculated to be 2.19 and 4.02 eV for thecalcinedTiO2 film and calcined keratin char-TiO2 composite film. The valence bands of the calcinedTiO2 film and calcined keratin char-TiO2 composite film were −7.90 and −7.70 eV with respect to vacuum level, respectively. Considering the band gaps obtained from Tauc's plot, the corresponding conduction band energies were −4.64 eV for the calcinedTiO2 film and −4.38 eV for the calcined keratin char-TiO2 composite film with respect to the vacuum level. As a result, the valence band and conduction band energies were respectively3.40 and 0.14 eV for thecalcinedTiO2 film and 3.20 and −0.12 eV for the calcined keratin char-TiO2 compositefilm versus the normal hydrogen electrode (NHE).

Separation Efficiency of Photogenerated Electron-Hole Pairs
The photocurrent-time curves and EIS plots of the calcined TiO2 film and calcined keratin char-TiO2 composite film are exhibited in Figure 8 to judge the separation and transfer efficiencies of photogenerated electron-hole pairs using the amperometric method. As the light turned on and off, the photocurrent responses were detected in both films which were derived from the photogenerated electron and hole separation. The photocurrent density of the calcined TiO2 film was larger than that of the calcined keratin char-TiO2 composite film. Hence, incorporating wool keratins into TiO2 might hinder the separation and transfer of photoinduced charge carriers in TiO2 [54]. As the light on/off cycle increased, the photocurrent intensity decreased slightly because of the recombination of photogenerated electron and hole pairs in the photo-response process [55].
The EIS Nyquist plots indicated that both the semicircle arcs of the calcined films were very close in the high-frequency region, implying there was no significant difference in the charge-transfer resistance for the calcined TiO2 film and calcined keratin char-TiO2 composite film [54]. However, in the low-frequency region, the calcined keratin char-TiO2 composite film's arc radius was larger than the calcined TiO2 film, suggesting the diffusion process of photogenerated charge carriers was inhibited in the calcined keratin char-TiO2 composite film. The impedance of the calcined keratin char-TiO2 composite film was The band gap energies of the TiO 2 films and keratin char-TiO 2 composite films were estimated to be 3.53 and 3.50 eV which were reduced to 3.26 and 3.32 eV after the calcination, respectively. The band gap of the calcined TiO 2 film was very close to the theoretic value 3.2 eV of anatase TiO 2 . The narrowed band gap was ascribed to the crystallization of TiO 2 under high-temperature conditions [20]. The C, N, and S co-doping could broaden the band gap of keratin char-TiO 2 composite film. However, the band gap of rutile TiO 2 modified wool powders was slightly reduced compared with the pure rutile TiO 2 particles [30]. The difference in the band gaps between the keratin char-TiO 2 composite film and TiO 2 film might be attributed to their different structures, crystallite sizes, electron mobility abilities, and activation energies [52].
The cutoff energies of the calcinedTiO 2 film and keratin char-TiO 2 composite film were separately estimated to be 15.49 and 17.52 eV by extrapolating the straight portion of the UPS curve at high binding energy region to zero. It was demonstrated that the work function of semiconductors could be calculated by subtracting the cutoff energy from the photon energy of the He I light source (21.2 eV) [53]. Thus, the work functions of thecalcinedTiO 2 film and calcined keratin char-TiO 2 composite film were 5.71 and 3.68 eV with respect to the vacuum level. The Fermi energies of thecalcinedTiO 2 film and calcined keratin char-TiO 2 composite film were −5.71 and −3.68 eV under vacuum conditions, respectively. Moreover, the valence band energy was estimated by extrapolating the straight line of the UPS curve at the low binding energy region to zero, which was lower than its Fermi energy. The valence band energies versus vacuum level were respectively calculated to be 2.19 and 4.02 eV for thecalcinedTiO 2 film and calcined keratin char-TiO 2 composite film. The valence bands of the calcinedTiO 2 film and calcined keratin char-TiO 2 composite film were −7.90 and −7.70 eV with respect to vacuum level, respectively. Considering the band gaps obtained from Tauc's plot, the corresponding conduction band energies were −4.64 eV for the calcinedTiO 2 film and −4.38 eV for the calcined keratin char-TiO 2 composite film with respect to the vacuum level. As a result, the valence band and conduction band energies were respectively3.40 and 0.14 eV for thecalcinedTiO 2 film and 3.20 and −0.12 eV for the calcined keratin char-TiO 2 compositefilm versus the normal hydrogen electrode (NHE).

Separation Efficiency of Photogenerated Electron-Hole Pairs
The photocurrent-time curves and EIS plots of the calcined TiO 2 film and calcined keratin char-TiO 2 composite film are exhibited in Figure 8 to judge the separation and transfer efficiencies of photogenerated electron-hole pairs using the amperometric method. As the light turned on and off, the photocurrent responses were detected in both films which were derived from the photogenerated electron and hole separation. The photocurrent density of the calcined TiO 2 film was larger than that of the calcined keratin char-TiO 2 composite film. Hence, incorporating wool keratins into TiO 2 might hinder the separation and transfer of photoinduced charge carriers in TiO 2 [54]. As the light on/off cycle increased, the photocurrent intensity decreased slightly because of the recombination of photogenerated electron and hole pairs in the photo-response process [55].

Separation Efficiency of Photogenerated Electron-Hole Pairs
The photocurrent-time curves and EIS plots of the calcined TiO2 film and calcined keratin char-TiO2 composite film are exhibited in Figure 8 to judge the separation and transfer efficiencies of photogenerated electron-hole pairs using the amperometric method. As the light turned on and off, the photocurrent responses were detected in both films which were derived from the photogenerated electron and hole separation. The photocurrent density of the calcined TiO2 film was larger than that of the calcined keratin char-TiO2 composite film. Hence, incorporating wool keratins into TiO2 might hinder the separation and transfer of photoinduced charge carriers in TiO2 [54]. As the light on/off cycle increased, the photocurrent intensity decreased slightly because of the recombination of photogenerated electron and hole pairs in the photo-response process [55].
The EIS Nyquist plots indicated that both the semicircle arcs of the calcined films were very close in the high-frequency region, implying there was no significant difference in the charge-transfer resistance for the calcined TiO2 film and calcined keratin char-TiO2 composite film [54]. However, in the low-frequency region, the calcined keratin char-TiO2 composite film's arc radius was larger than the calcined TiO2 film, suggesting the diffusion process of photogenerated charge carriers was inhibited in the calcined keratin char-TiO2 composite film. The impedance of the calcined keratin char-TiO2 composite film was larger than that of the calcined TiO2 film. This was probably attributed to the fact that the calcined keratin char-TiO2 composite film had the elements of N and S, although the C content (20.89% in atomic conc.) of the calcined keratin char-TiO2 composite film was slightly larger than that (19.56% in atomic conc.) of the calcined TiO2 film.   The EIS Nyquist plots indicated that both the semicircle arcs of the calcined films were very close in the high-frequency region, implying there was no significant difference in the charge-transfer resistance for the calcined TiO 2 film and calcined keratin char-TiO 2 composite film [54]. However, in the low-frequency region, the calcined keratin char-TiO 2 composite film's arc radius was larger than the calcined TiO 2 film, suggesting the diffusion process of photogenerated charge carriers was inhibited in the calcined keratin char-TiO 2 composite film. The impedance of the calcined keratin char-TiO 2 composite film was larger than that of the calcined TiO 2 film. This was probably attributed to the fact that the calcined keratin char-TiO 2 composite film had the elements of N and S, although the C content (20.89% in atomic conc.) of the calcined keratin char-TiO 2 composite film was slightly larger than that (19.56% in atomic conc.) of the calcined TiO 2 film.

Photocatalysis Performance
The adsorption ability (in the dark), photocatalytic degradation, cycle stability, and trapping experiments of MB dye solutions by the TiO 2 films and keratin char-TiO 2 composite films under visible light irradiation are described in Figure 9. It was evident that after 1 h of adsorption in the dark, the removal rate of MB dye from the aqueous solution by the keratin char-TiO 2 film (29.9%) was much larger than that of the TiO 2 film (13.6%). The removal rate of MB dye by the calcined keratin char-TiO 2 composite film was reduced to 18.9% but larger than that of the calcined TiO 2 film (14.1%). As compared with the calcined TiO 2 film, the strong adsorption ability of the calcined keratin char-TiO 2 composite film towards the MB dye was verified by the following DFT theoretical calculations. The k value for the blank MB dye solution was calculated to be 1.39 × 10 −3 min −1 (R 2 = 0.995), indicating the visible light had a little effect on the photodegradation of the MB due to the photosensitization of MB molecules [56]. With respect to the uncalcined films, when wool keratin was doped with TiO 2 , after 180 min visible light illumination, the degradation rate and k value of the keratin char-TiO 2 composite films increased from 51.4% and 4.01 × 10 −3 min −1 (R 2 = 0.999) to 80.6% and 8.90 × 10 −3 min −1 (R 2 = 0.993), respectively. This was probably because the wool keratin sulfhydryl group (-SH-) had a strong complexing property and promoted the contact between TiO 2 and MB molecules [57]. After the films were calcinated, the k values of the calcined TiO 2 film and calcined keratin char-TiO 2 composite film separately increased to 7.67 × 10 −3 min −1 (R 2 = 0.999) and 11.22 × 10 −3 min −1 (R 2 = 0.998). It meant that the high-temperature treatment resulted in the incorporation of C, N, and S elements of wool keratin into TiO 2 lattice and the crystallization of TiO 2 as confirmed by XPS and XRD results. Thus the electron energy band structures and the hydrophilic abilities of the calcined TiO 2 film and calcined keratin char-TiO 2 composite film were changed. The separation and transfer of photogenerated electron and hole pairs were improved accordingly. It was confirmed that the non-metal element doping could introduce the O 2 sites in TiO 2 to mediate its valence band and reduce its band gap. At the same time, the catalyst density, the specific surface area, the concentration of valence surface oxygen, and the surface reaction sites increased to some extent, which further promoted the photocatalytic activity of TiO 2 [12].
The photodegradation stability results indicated that after five successive MB photodegradation, the degradation rates decreased slightly from 75.7% to 73.4% for the calcined TiO 2 film and from 87.4% to 83.6% for the calcined keratin char-TiO 2 composite film. The long-term light irradiation might degrade the C-C bonds, decompose the protein compositions, and increase the carbon defects in the calcined keratin char-TiO 2 composite film during the photocatalytic process [58]. Hence, the photocatalytic activities of both the calcined films were well maintained for the photodegradation of the MB solution under visible light irradiation.
To detect the reactive species of h + , 1 O 2 , •OH, and •O 2 − , the trapping experiments of the calcined films were conducted by adding the scavengers of EDTA, FA, TBA, and BQ in MB solutions under visible light, respectively. Obviously, after being added with EDTA, FA, TBA, and BQ individually, the k values decreased to different extents. Both the calcined films follow the same order: FA > EDTA > BQ > TBA, and thereby 1 O 2 played a decisive role in the MB photodegradation process. The k values were reduced to 1.19 × 10 −3 min −1 (R 2 = 0.954) and 0.81 × 10 −3 min −1 (R 2 = 0.857) for the calcined TiO 2 film and calcined keratin char-TiO 2 composite film, respectively. At the same time, h + , •O 2 − and •OH also affected the MB photodegradation process. The reactive order of radical species was not in agreement with the previous studies [31]. Therefore, the doping of wool keratin had no influence on the radical species of calcined TiO 2 film in the MB photodegradation process under visible light irradiation.

DFT Theoretical Calculations
The schematic constructions of adsorption energy and energy band structures for the calcined TiO2 film and calcined keratinchar-TiO2 composite film were simulated using the DFT calculations, as displayed in Figure 10. As shown in Figure 10a,b, the relaxed atom structures of an MB molecule were separately adsorbed on the calcined TiO2 (C-doped TiO2) and calcined keratin char-TiO2 (C, N and S co-doped TiO2). According to Equation (5), the Eads values were calculated to be −0.514 eV for the C-doped TiO2 and −0.891 eV for the C, N, and S co-doped TiO2 [59]. The higher Eadsof−0.891 eV implied that the adsorption of the MB molecules on the calcined keratin char-TiO2 composite film occurred more easily. Thus, the co-doping of C, N, and S elements of keratin into anatase TiO2 was a benefit for the adsorption of MB on the calcined keratin char-TiO2 composite film. This was a vital process for photocatalytic degradation. Although the calcined keratin char-TiO2 compo-

DFT Theoretical Calculations
The schematic constructions of adsorption energy and energy band structures for the calcined TiO 2 film and calcined keratinchar-TiO 2 composite film were simulated using the DFT calculations, as displayed in Figure 10. As shown in Figure 10a,b, the relaxed atom structures of an MB molecule were separately adsorbed on the calcined TiO 2 (Cdoped TiO 2 ) and calcined keratin char-TiO 2 (C, N and S co-doped TiO 2 ). According to Equation (5), the E ads values were calculated to be −0.514 eV for the C-doped TiO 2 and −0.891 eV for the C, N, and S co-doped TiO 2 [59]. The higher E ads of −0.891 eV implied that the adsorption of the MB molecules on the calcined keratin char-TiO 2 composite film occurred more easily. Thus, the co-doping of C, N, and S elements of keratin into anatase TiO 2 was a benefit for the adsorption of MB on the calcined keratin char-TiO 2 composite film. This was a vital process for photocatalytic degradation. Although the calcined keratin char-TiO 2 composite film's band gap increased slightly compared to the calcined TiO 2 film, the photocatalytic degradation performance of the calcined keratin char-TiO 2 composite film did not become deteriorated. The strong adsorption of MB on the composite film played a privileged role in the MB photocatalytic degradation process [60].
Catalysts 2021, 11,1366 14 of 23 site film's band gap increased slightly compared to the calcined TiO2 film, the photocatalytic degradation performance of the calcined keratin char-TiO2 composite film did not become deteriorated. The strong adsorption of MB on the composite film played a privileged role in the MB photocatalytic degradation process [60]. The energy band structures along the high symmetry directions were calculated to investigate the underlying electronic properties of calcined TiO2 film and calcined keratin char-TiO2 composite film. As shown in Figure 10c,d, because the calculated conduction band minima and valence band maxima for both the calcined films were located at different high symmetry points, they had indirect band gaps. Generally, the photogenerated charge carriers cannot recombine directly for the indirect band gap semiconductors, which is favorable for the photocatalytic reaction [61]. The calculated band gaps were 3.11 and 3.25 eV for the calcined TiO2 film and calcined keratin char-TiO2 composite film, which were slightly smaller than the experimental results as shown in Figure 7d (3.26 eV for the calcined TiO2 film and 3.32 eV for the calcined keratin char-TiO2 composite film), respectively. Therefore, it was concluded that the band gap of the calcined keratin char-TiO2 composite film increased when the C, N, and S elements of wool keratins were doped into anatase TiO2.

Hydrophobic Behaviors
The time-sequence photographs of water droplets (Congo red was used for better showing) sliding process on the surfaces of the calcined TiO2 film and calcined keratinchar-TiO2 composite film at different sliding angles were shown in Figure 11. It was found that no matter what the film tilt angle was, the water droplets could spread out and slip off on the calcined TiO2 film leaving their marks (Figure 11a,b). On the contrary, the water droplets were adhered to the surface of the calcined keratinchar-TiO2 composite film and were about to roll off when the film tilt angle was smaller than 30°.Once the inclination The energy band structures along the high symmetry directions were calculated to investigate the underlying electronic properties of calcined TiO 2 film and calcined keratin char-TiO 2 composite film. As shown in Figure 10c,d, because the calculated conduction band minima and valence band maxima for both the calcined films were located at different high symmetry points, they had indirect band gaps. Generally, the photogenerated charge carriers cannot recombine directly for the indirect band gap semiconductors, which is favorable for the photocatalytic reaction [61]. The calculated band gaps were 3.11 and 3.25 eV for the calcined TiO 2 film and calcined keratin char-TiO 2 composite film, which were slightly smaller than the experimental results as shown in Figure 7d (3.26 eV for the calcined TiO 2 film and 3.32 eV for the calcined keratin char-TiO 2 composite film), respectively. Therefore, it was concluded that the band gap of the calcined keratin char-TiO 2 composite film increased when the C, N, and S elements of wool keratins were doped into anatase TiO 2 .

Hydrophobic Behaviors
The time-sequence photographs of water droplets (Congo red was used for better showing) sliding process on the surfaces of the calcined TiO 2 film and calcined keratinchar-TiO 2 composite film at different sliding angles were shown in Figure 11. It was found that no matter what the film tilt angle was, the water droplets could spread out and slip off on the calcined TiO 2 film leaving their marks (Figure 11a,b). On the contrary, the water droplets were adhered to the surface of the calcined keratinchar-TiO 2 composite film and were about to roll off when the film tilt angle was smaller than 30 • .Once the inclination angle of the composite film was larger than 45 • , the water droplets quickly rolled off from its surface without any residual liquid left. This was due to the large water contact angle and rough surface of the calcined keratin char-TiO 2 composite film [62]. Hence, the C, N, and S elements co-doping with TiO 2 could significantly improve the hydrophobicity of the calcined keratinchar-TiO 2 composite film.
It is known that the hydrophobic self-cleaning film cannot be resistant to acid and alkali corrosion, resulting in poor waterproof and decontamination performance [63]. Here, the hydrophobicity of the films was further assessed by acid and base immersion tests, as illustrated in Figure 12. The calcined TiO2 film and calcined keratinchar-TiO2 composite film were separately immersed in 10% HCl and 10% alkaline aqueous solution at 30 °C for 24 h and then washed with deionized water to remove the residual chemicals. No change of water droplets state was observed on both acid-and alkali-treated keratinchar-TiO2 composite films, exhibiting better corrosion resistance. However, the hydrophilicity of the alkali-treated TiO2 film was improved to some degree. This was caused by the increased surface free energy resultant from OH − groups and small water contact angle [64].  It is known that the hydrophobic self-cleaning film cannot be resistant to acid and alkali corrosion, resulting in poor waterproof and decontamination performance [63]. Here, the hydrophobicity of the films was further assessed by acid and base immersion tests, as illustrated in Figure 12. The calcined TiO 2 film and calcined keratinchar-TiO 2 composite film were separately immersed in 10% HCl and 10% alkaline aqueous solution at 30 • C for 24 h and then washed with deionized water to remove the residual chemicals. No change of water droplets state was observed on both acid-and alkali-treated keratinchar-TiO 2 composite films, exhibiting better corrosion resistance. However, the hydrophilicity of the alkali-treated TiO 2 film was improved to some degree. This was caused by the increased surface free energy resultant from OH − groups and small water contact angle [64].

Photocatalytic Reaction Mechanism
Based on the above characterization analysis and DFT calculations, the photo mechanisms for the MB photodegradation by the calcined TiO2 film and calcined char-TiO2 composite film are explicated in Figure 13. The atom mass ratios (in pe C1s to Ti2p were 19.56:24.34 for the calcined TiO2 film and 22.89:23.00 for the calci atin char-TiO2 composite film. Thus, the amount of keratin char layer was less amount of TiO2;a certain amount of TiO2 nanoparticles could be exposed to the ai films. For both the calcined films, the valence band potentials were more positive redox potential of OH − /OH (1.99 eV vs. NHE)couples, meanwhile, the conduct potentials were more negative than the redox potential of O2/O2 − (−0.33 eV vs. N ples. Thus, the photogenerated holes could oxidize H2O to generateOH radical while, the photogenerated electrons could reduce dissolved O2 to produceO2 − rad cause the absolute value of the adsorption energy (−0.891 eV) of calcined keratin c composite film was larger than that (−0.514 eV) of the calcined TiO2 film, the MB ecules were closely adsorbed on the calcined keratin char-TiO2 composite film. T toinduced active species like 1 O2, h + , •O2 − and •OH could easily/quickly react with molecules. Thus, the calcined keratinchar-TiO2 composite film containing N and result in enhanced photocatalytic activities. Due to the presence of carbon laye films, as demonstrated by TEM and XPS analyses, the photogenerated electro prone to transfer to the carbon layer. In contrast, the photogenerated holes would TiO2.Hence, the recombination of photogenerated electron and hole pairs in both cined films could be effectively inhibited. As confirmed by the trapping experim photogenerated electrons could react with dissolved O2 to produce •O2 − radical further reacted with the photogenerated h + to generate 1 O2 radicals. The formatio 1 O2 radical was faster than that of OH radical because the reaction route of 1 O2 wa than that of OH. Thus, the main active species was 1 O2 radicals in the MB photod tion process. In addition, the calcined TiO2 film and calcined keratin char-TiO2 co film had many meso-and micro-pores as confirmed by FESEM, which might faci adsorption of MB dyes molecules. Hence, the adsorption capacities for both film benefit for the removal of dyes from the solution.

Photocatalytic Reaction Mechanism
Based on the above characterization analysis and DFT calculations, the photocatalytic mechanisms for the MB photodegradation by the calcined TiO 2 film and calcined keratin char-TiO 2 composite film are explicated in Figure 13. The atom mass ratios (in percent) of C 1s to Ti 2p were 19.56:24.34 for the calcined TiO 2 film and 22.89:23.00 for the calcined keratin char-TiO 2 composite film. Thus, the amount of keratin char layer was less than the amount of TiO 2 ; a certain amount of TiO 2 nanoparticles could be exposed to the air in both films. For both the calcined films, the valence band potentials were more positive than the redox potential of OH − /•OH (1.99 eV vs. NHE) couples, meanwhile, the conduction band potentials were more negative than the redox potential of O 2 /•O 2 − (−0.33 eV vs. NHE) couples. Thus, the photogenerated holes could oxidize H 2 O to generate OH radicals, meanwhile, the photogenerated electrons could reduce dissolved O 2 to produce O 2 − radicals. Because the absolute value of the adsorption energy (−0.891 eV) of calcined keratin char-TiO 2 composite film was larger than that (−0.514 eV) of the calcined TiO 2 film, the MB dye molecules were closely adsorbed on the calcined keratin char-TiO 2 composite film. The photoinduced active species like 1 O 2 , h + , •O 2 − and •OH could easily/quickly react with the dye molecules. Thus, the calcined keratinchar-TiO 2 composite film containing N and S could result in enhanced photocatalytic activities. Due to the presence of carbon layer in both films, as demonstrated by TEM and XPS analyses, the photogenerated electrons were prone to transfer to the carbon layer. In contrast, the photogenerated holes would stay on TiO 2 .Hence, the recombination of photogenerated electron and hole pairs in both the calcined films could be effectively inhibited. As confirmed by the trapping experiments, the photogenerated electrons could react with dissolved O 2 to produce •O 2 − radicals, which further reacted with the photogenerated h + to generate 1 O 2 radicals. The formation rate of 1 O 2 radical was faster than that of OH radical because the reaction route of 1 O 2 was shorter than that of OH. Thus, the main active species was 1 O 2 radicals in the MB photodegradation process. In addition, the calcined TiO 2 film and calcined keratin char-TiO 2 composite film had many meso-and micro-pores as confirmed by FESEM, which might facilitate the adsorption of MB dyes molecules. Hence, the adsorption capacities for both films were a benefit for the removal of dyes from the solution.

Preparation of Keratin char-TiO2 Composite Films
The procedure for the preparation of keratin char-TiO2 composite films is shown in Figure 14. Firstly, wool keratins were obtained from cleaned wool fibers using the chemical reduction technique. Next, the wool keratin-TiO2 composite films were prepared by alternatively spin-coating of keratin aqueous solution and TiO2 precursor solution on quartz glass slides. Lastly, the wool keratin-TiO2 composite films were calcined at 400 °C in an air environment to obtain the keratin char-TiO2 composite films.

Materials and Reagents
Wool fibers were obtained from a local textile mill, and the fineness was in the range from 21.

Preparation of Keratin char-TiO 2 Composite Films
The procedure for the preparation of keratin char-TiO 2 composite films is shown in Figure 14. Firstly, wool keratins were obtained from cleaned wool fibers using the chemical reduction technique. Next, the wool keratin-TiO 2 composite films were prepared by alternatively spin-coating of keratin aqueous solution and TiO 2 precursor solution on quartz glass slides. Lastly, the wool keratin-TiO 2 composite films were calcined at 400 • C in an air environment to obtain the keratin char-TiO 2 composite films.

Preparation of Keratin char-TiO2 Composite Films
The procedure for the preparation of keratin char-TiO2 composite films is shown in Figure 14. Firstly, wool keratins were obtained from cleaned wool fibers using the chemical reduction technique. Next, the wool keratin-TiO2 composite films were prepared by alternatively spin-coating of keratin aqueous solution and TiO2 precursor solution on quartz glass slides. Lastly, the wool keratin-TiO2 composite films were calcined at 400 °C in an air environment to obtain the keratin char-TiO2 composite films.

Extraction of Wool Keratins
The as-obtained wool fibers were first degreased on a Soxhlet apparatus using petroleum ether as the solvent to remove the lipids and impurities of wools at 80 • C for 12 h [65], then successive washed with absolute ethanol and deionized water, and lastly dried at 60 • C overnight. The degreased wool fibers (0.1 g) were cut into snippets with less than 5 mm length and then immersed in 10 mL of mixture solution containing0.5 M Na 2 S 2 O 5 , 8 M urea, and 0.2 M SDS based on a chemical reduction method [66]. The mixture solution was heated to 100 • C and treated at a constant temperature for 2 h under continuous magnetic stirring. After that, 40 mL deionized water was poured into the mixture solution, filtered with a400-mesh stainless-steel mesh screen with a diameter of 0.037 mm to remove the insoluble substances. The filtrate was dialyzed against deionized water using a cellulose tube with a cutoff of 3500 Da for 3 days at room temperature. During the dialysis procedure, the deionized water was changed every 6 h. The resultant aqueous solution was condensed with a lyophilizer to obtain the pristine wool keratin powders.

Spin-Coating and Calcinations of Wool Keratin-TiO 2 Composite Films
The wool keratin-TiO 2 composite films were deposited on quartz glass slides based on a sol-gel spin-coating technique [67]. Prior to spin-coating, the glass slides were cleaned with piranha solution (H 2 O 2 :H 2 SO 4 = 1:3 in volumes) to remove any organic contaminants and native oxide layer on the substrate [68]. A thin layer of APES was first spin-coated on the clear glass slide with 2 vol% APES absolute ethanol solution to improve the binding ability between wool keratin and glass substrate. Next, 0.142 g TTIP and 0.2 mL acetic acid were slowly added into 100 mL isopropyl alcohol under stirring to obtain the TTIP precursor solution. Meanwhile, 3 mg of wool keratin powders was added in 10 mL deionized water to obtain the keratin solution, which was spin-coated on the glass slide at a speed of 4000 rpm for 30 s. The glass slide was immediately heated at 80 • C for 1 min. The TTIP precursor solution was subsequently spin-coated on the keratin-coated slide at 4000 rpm for 30 s twice and then treated at 100 • C for 10 min. The quartz glass slide was alternately spin-coated with wool keratin and TTIP precursor solutions for 50 cycles to obtain the wool keratin-TiO 2 composite films. The TiO 2 films were also fabricated by the spin-coating method only with the TTIP precursor solution. Lastly, the spin-coated quartz glass slides were subjected to annealing at 400 • C for 1 h and cooled down naturally.

Characterization Methods
According to the Bradford protein assay method [69], the molecular weight distribution of as-prepared wool keratins was determined on the Mini-Protean system by the SDS-polyacrylamide gel electrophoresis (PAGE). The wool keratin powders were dissolved in a buffer solution containing Tris/HCl (1.25 mM, pH 6. The morphologies of as-prepared films were examined on a Quanta-450-FEG+X-MAX50 field emission scanning electron microscope (FESEM, FEI Company, Hillsboro, OR, USA). The specimens were sprayed with conductive gold coating under vacuum prior to testing.
The surface roughness of the films was investigated on a Laser electron microscope (Carl Zeiss AG, Oberkochen, Germany).
The crystal structures of the films were analyzed on a 7000S X-ray diffractometer (Shimazu Corp., Kyoto, Japan) operated at 40 kV and 30 mA at a scanning rate of 8 • /min. The Cu Kα 1 X-ray source (λ = 0.154056 nm) was used. The crystallinity size (D) was calculated according to the Scherrer Equation (1) [70] as below.
where k, λ, β and θ were a constant (0.89), X-ray wavelength, full width at half maximum (FWHM), and Bragg angle, respectively. The microstructures of the calcined films were observed on a JEM-3010 transmission electron microscope (TEM, JEOL Ltd., Tokyo, Japan).
The static water contact angleof the films was conducted on an OCA25 contact angle meter (Data Physics Instruments, Filderstadt, Germany) according to the drop method.
The chemical bonding states of the films were carried out on an Axis Ultra electron spectrometer (Kratos Analytical Ltd., Manchester, UK) operated at 10 kV and 10 mA at the vacuum below 10 −8 Torr. The monochromatic Al Kα (1486.68 eV) radiation was used, and the binding energies were calibrated through the 284.8 eV C 1s peak.
The diffuse reflectance spectra of the films were recorded on a Lambda 950 spectrophotometer (PerkinElmer Inc., Waltham, MA, USA) equipped with a 150 mm integrating sphere. The barium sulfate was applied to calibrate the 100% reflectance. The band gap energy (E g ) was deduced from the intersection of the extrapolation of (αhv) 1/2 versus hv according to the Tauc Equation (2) [38] as below: where α was the absorption coefficient, ν was the light frequency, h was the Planck constant, A was a constant which was determined by the nature of allowed transition (n = 1/2 for direct and 2 for indirect). The calcined films' ultraviolet photoelectron spectroscopy (UPS) was performed on the Escalab 250 Xi UV photoelectron spectroscope (Thermo Fisher Scientific Inc., Waltham, MA, USA).
The photocurrent response curves and electrochemical impedance spectroscopy (EIS) Nyquist plots of the calcined films were measured on a CHI760E electrochemical workstation (Shanghai Chenhua Instrument Co., Ltd. Shanghai, China) based on the three-electrode system. A 300 W Xenon lamp was used to simulate visible lights along with a 420 nm cutoff filter. The Pt foil and saturated calomel electrode (SCE) were used as the counter and reference electrodes. The working electrode was prepared by spin-coating wool keratin and/or TTIP precursor solutions on an indium-tin-oxide (ITO) glass substrate with a size of 2 × 1 cm 2 . The photocurrent-time curves were collected under visible lights by turning on/off cycles at 0.5 V vs. SCE. The electrolyte was 0.5 M Na 2 SO 4 aqueous solution at the pH value of 7, and the experimental equipment ran for 30 s firstly, and then each illumination time was 30 s. The frequency for EIS measurements was from 0.1 to 100,000 Hz at an alternating current voltage amplitude of 5 mV under open-circuit conditions.

Measurements of Photocatalytic Properties
The photocatalytic properties of the as-prepared films were assessed by measuring the photodegradation of model MB dye under visible light irradiation. In a typical experiment, the film specimen (4 × 4 cm 2 ) was put in a glass petri dish (φ = 90 mm) containing 50 mL of 1.0 mg/L MB solution. The pH of the MB solution was 6.8. Prior to illumination, the dye solution with the film was placed in the dark for 60 min to reach an adsorption−desorption equilibrium. After that, a 30 W LED lamp was applied on the dye solution, and the illuminance of the light was 45,000 Lux. At a given time interval t, 3 mL solution was extracted from the dye solution. The absorbance A t was recorded on a VIS-7220 visible light spectroscope (Beijing Beifen-Ruili Analytical Instrument Co. Ltd., Beijing, China) at 664 nm. The concentration C t was calculated according to the MB calibration curve (A t = 0.0107 + 0.1767C t , R 2 = 0.99). The degradation rate D (%) was calculated according to Equation (3) as below: where C 0 and C t were respectively the MB concentration at the initial time and time t. To quantitatively identify the photodegradation of MB solution, the reaction kinetic data were fitted based on the Langmuir-Hinshelwood model [71]. The pseudo-first-order apparent rate constant k (min −1 ) was calculated by Equation (4) as below: In addition, the oxidized MB dye solution was replaced by the fresh MB solution, and the cycle photodegradation of MB solution was measured under the same visible light irradiation condition. The active species of h + , 1 O 2 , •OH, and •O 2 − in the photocatalytic solutions were identified by adding 1.0 mM EDTA, 1.0 mL FA [72], 0.5 mL TBA, and 0.2 mM BQ [73] as the scavengers, respectively.

Theoretical Calculation Method
The DFT calculations were implemented using the Vienna Ab Initio Package [74,75] program based on the Perdew-Burke-Ernzerhof generalized gradient approximation [76]. The ionic cores were described by the projected augmented wave potentials [77,78], and the valence electrons were considered. For the plane-wave basis set, an energy cutoff of 450 eV was used. The Gaussian smearing techniquewith0.05 eV width was applied to optimize the partial occupancies of the Kohn−Sham orbitals. Provided that the energy change was less than 10−4 eV, self-consistent electronic energy was realized. When the force change was below 0.05 eV/Å, the geometry optimization was supposed to be convergent. The empirical dispersion correction methodology [79] was employed to explicate the dispersion interactions.
The Brillouin zone sampling with a 10 × 10 × 4 Monkhorst-Pack k-point grid was used to optimize the equilibrium lattice constants of TiO 2 unit cell. A super cell Graphene model was constructed in the x and y directions. The 3 × 3 × 1 Monkhorst-Pack k-point grid was used for the Brillouin zone sampling for the C, N and S atoms doping in the super cell TiO 2 structure in the structural optimization process. Thus, the adsorption energy(E ads ) could be calculated according to Equation (5) as below: where E ad/sub was the total energies of the optimized adsorbate/substrate system, E ad was the adsorbate energy of the structure, and E sub was the energy of the clean substrate.

Conclusions
The hydrophobic self-cleaning keratin char-TiO 2 composite films were made from the calcination of keratin-TiO 2 films for the photocatalytic degradation of dye pollutants. The incorporation of TiO 2 precursors in the wool keratin film led to the formation of anatase TiO 2 nanoparticles after calcination. The water contact angle and band gap of the calcined keratin char-TiO 2 composite film were greater than the calcinated TiO 2 film. Although the separation of photogenerated charge carriers was inhibited, the calcined keratin char-TiO 2 composite film exhibited better photocatalytic performance in the MB degradation process and good hydrophobic behaviors to shield off water droplets. Both experiments and DFT calculations demonstrated that the high photocatalytic properties of the calcined keratin char-TiO 2 composite film were mainly ascribed to its rough surface, particular porous structure, large water contact angle, and high adsorption energy, which were helpful for the contact of organic compounds with the composite film. 1 O 2 radicals were the dominant reactive species. Overall, the calcined keratin char-TiO 2 composite film obtained could be used in the photocatalysis field.
Author Contributions: H.Z. conceptualized, designed, and administered the experiments. J.Z. contributed to the synthesis and characterization of materials. W.L., L.Y. and H.W. helped in collecting and analyzing data. N.M. helped to revise the grammar and gave conceptual advice. All authors analyzed and discussed the data. J.Z. and H.Z. wrote the manuscript. All authors have read and agreed to the published version of the manuscript.

Conflicts of Interest:
The authors declare no conflict of interest.