C-,N- and S-Doped TiO2 Photocatalysts: A Review

This article presents an overview of the reports on the doping of TiO2 with carbon, nitrogen, and sulfur, including single, co-, and tri-doping. A comparison of the properties of the photocatalysts synthesized from various precursors of TiO2 and C, N, or S dopants is summarized. Selected methods of synthesis of the non-metal doped TiO2 are also described. Furthermore, the influence of the preparation conditions on the doping mode (interstitial or substitutional) with reference to various types of the modified TiO2 is summarized. The mechanisms of photocatalysis for the different modes of the non-metal doping are also discussed. Moreover, selected applications of the non-metal doped TiO2 photocatalysts are shown, including the removal of organic compounds from water/wastewater, air purification, production of hydrogen, lithium storage, inactivation of bacteria, or carbon dioxide reduction.


Introduction
More than ever before, environmental problems have become a major concern. Urbanization and rapid growth of industries generate abundant amounts of pollutants which are released into the environment. Among them, there are highly hazardous materials such as pharmaceuticals [1,2], dioxins [3], pesticides [4], herbicides [1], phenols [4,5], and textile dyes [1,[4][5][6]. This increasing occurrence of organic pollutants in the environment is a serious danger for health and the lives of humans and other living beings. Conventional treatment methods very often fail in the removal of these kinds of residues entirely, because of their high (bio)chemical stability. Moreover, a conventional approach is associated with the operational problems and high costs. Hence, the development of new and efficient methods of the removal of organic contaminants is a matter of growing interest [1,2,4].
In recent decades, semiconductor photocatalysis has been proved to be an efficient approach for organic compounds decomposition and degradation. TiO 2 has been widely and successfully used as a photocatalyst in many different areas ( Figure 1) due to its advantages, such as low cost and good chemical stability. However, it requires employing relatively high photon energy to be activated. For this reason, many methods of narrowing of the band gap of TiO 2 have been proposed, aimed at the direct usage of sunlight [7,8]. Amongst them, doping of TiO 2 with non-metals such as carbon, nitrogen and sulfur is often reported as one of the most effective ways of increasing its photocatalytic activity under visible light [1,9,10]. Non-metal doping of TiO 2 leads to changes in the electronic band structure, resulting in a smaller band gap energy value, and thus an improved response in the visible light [7,8,11]. The idea of non-metal doping of TiO2 has been discussed in numerous reviews through the years [1,9,10,[12][13][14][15][16][17][18][19][20][21][22][23]. Most of the recent reviews referred to nitrogen only, which is one of the most frequently used non-metal dopants [17,[24][25][26][27][28][29][30][31][32][33][34]. On the other hand, the reviews devoted exclusively to S-doped TiO2 are very limited [35]. In some papers, the modifications of TiO2 with carbon are also summarized [1,10,36,37]. Shi et el. [36] presented various carbon-based (nano)composites, including C-doped TiO2. Moreover, diverse, more complex configurations, e.g., with multi-walled carbon nanotubes (MWCNT) in TiO2-SiO2/MWCNT [37], with carbon dots (CDs) in CDs-N-TiO2 [38], and Ag-modified g-C3N4/N-doped TiO2 [24] have been reported. There is only one review referring to the beneficial effects and challenges of tri-doping of TiO2 with carbon, nitrogen, and sulfur, which was published in 2017 [39]. Moreover, recently, a review on single doping of TiO2 with various non-metals, including C, N, and S was published [1].
In contrast with the above papers, this review focuses on a complex comparison of various types and configurations of TiO2 doping with non-metals such as carbon, nitrogen, and sulfur, including single-, co-, and tri-doping. The aim of this work was to systematize the knowledge with reference to the abovementioned doping types, especially in terms of the doping mode (i.e., the way the dopant interacts with the unit cell of crystal structure or the surface of a semiconductor), the effect of the single-, co-, and tri-doping on the mechanism of photocatalysis, as well as the possible applications of the doped TiO2. The overview of the single C, N, and S doping creates a base for a better understanding of the co-and tri-doping. The first part of the review presents a brief introduction to the TiO2 photocatalysis. The influence of various non-metal dopants (C, N, and S) on the photocatalysts' properties is reported in the subsequent sections. A discussion on the type of the doping source, TiO2 precursor, and the photocatalyst preparation method is presented for all modes of doping of TiO2 with C, N, or S. The morphological and structural characteristics of the photocatalysts as well as the photocatalytic activity are also summarized. The literature reports selected for this review are distributed appropriately among Sections 3-9 according to the description of the doping type proposed by the authors of the cited papers.
In contrast with the above papers, this review focuses on a complex comparison of various types and configurations of TiO 2 doping with non-metals such as carbon, nitrogen, and sulfur, including single-, co-, and tri-doping. The aim of this work was to systematize the knowledge with reference to the abovementioned doping types, especially in terms of the doping mode (i.e., the way the dopant interacts with the unit cell of crystal structure or the surface of a semiconductor), the effect of the single-, co-, and tri-doping on the mechanism of photocatalysis, as well as the possible applications of the doped TiO 2 . The overview of the single C, N, and S doping creates a base for a better understanding of the co-and tri-doping. The first part of the review presents a brief introduction to the TiO 2 photocatalysis. The influence of various non-metal dopants (C, N, and S) on the photocatalysts' properties is reported in the subsequent sections. A discussion on the type of the doping source, TiO 2 precursor, and the photocatalyst preparation method is presented for all modes of doping of TiO 2 with C, N, or S. The morphological and structural characteristics of the photocatalysts as well as the photocatalytic activity are also summarized. The literature reports selected for this review are distributed appropriately among Sections 3-9 according to the description of the doping type proposed by the authors of the cited papers.

TiO 2 Photocatalysis
The discovery of the photocatalytic splitting of water on TiO 2 electrodes in 1972 heralded a new era of heterogeneous photocatalysis [40]. Despite several decades having passed since then, the most popular photocatalyst is still TiO 2 . Amongst the different structures of titania, anatase and rutile are commonly used in photocatalysis, with anatase displaying a higher photocatalytic activity.
Catalysts 2021, 11, 144 3 of 56 In heterogeneous photocatalysis, a reaction takes place on the surface of a photocatalyst. The general mechanism of photocatalytic decomposition of organic compounds is summarized in Figure 2.
Catalysts 2021, 11, x FOR PEER REVIEW 3 of 58 In heterogeneous photocatalysis, a reaction takes place on the surface of a photocatalyst. The general mechanism of photocatalytic decomposition of organic compounds is summarized in Figure 2. First, the energy higher than the band gap energy of the semiconductor is required for photon absorption and excitation of an electron (e − ) from the valence band (VB) to the conduction band (CB), resulting in hole (h + ) generation. The holes in the VB can react with the surface adsorbed water or hydroxyl ions to form hydroxyl radicals, which are extremely strong oxidants (oxidation potential around +2.7 V). The photoexcited electrons in the CB can generate superoxide radicals due to the reaction with oxygen, being the main electron acceptor in the system. Further reactions lead to the formation of other reactive oxygen species (ROS) such as hydrogen peroxide, hydroperoxyl radicals, or hydroxyl radicals. These species participate in the degradation of organic contaminants [41][42][43][44].
The effectiveness of the photodegradation of pollutants on the semiconductor surface is influenced by: (i) the chemical composition, structure and concentration of pollutants, (ii) the radiation intensity, (iii) the exposure time, (iv) the amount of photocatalyst used, (v) the oxygen content in the reaction medium, (vi) the pH value of the solution, (vii) the properties of a photocatalyst (specific surface area, crystallographic structure, number of surface defects, presence of additives and dopants, etc.) [45,46].
Photocatalysis can be also applied for the oxidation of inorganic compounds, such as nitrogen oxides. The mechanism of photocatalytic NOx oxidation under UV illumination is represented by Equations (1)-(7) [47]. The reactants are adsorbed on the photocatalyst surface: TiO + NO ↔ TiO NO TiO + NO ↔ TiO NO (4) and subsequently the NOx species undergo oxidation by the • OH radicals as follows: First, the energy higher than the band gap energy of the semiconductor is required for photon absorption and excitation of an electron (e − ) from the valence band (VB) to the conduction band (CB), resulting in hole (h + ) generation. The holes in the VB can react with the surface adsorbed water or hydroxyl ions to form hydroxyl radicals, which are extremely strong oxidants (oxidation potential around +2.7 V). The photoexcited electrons in the CB can generate superoxide radicals due to the reaction with oxygen, being the main electron acceptor in the system. Further reactions lead to the formation of other reactive oxygen species (ROS) such as hydrogen peroxide, hydroperoxyl radicals, or hydroxyl radicals. These species participate in the degradation of organic contaminants [41][42][43][44].
The effectiveness of the photodegradation of pollutants on the semiconductor surface is influenced by: (i) the chemical composition, structure and concentration of pollutants, (ii) the radiation intensity, (iii) the exposure time, (iv) the amount of photocatalyst used, (v) the oxygen content in the reaction medium, (vi) the pH value of the solution, (vii) the properties of a photocatalyst (specific surface area, crystallographic structure, number of surface defects, presence of additives and dopants, etc.) [45,46].
Photocatalysis can be also applied for the oxidation of inorganic compounds, such as nitrogen oxides. The mechanism of photocatalytic NO x oxidation under UV illumination is represented by Equations (1)-(7) [47]. The reactants are adsorbed on the photocatalyst surface: TiO 2 + NO 2 ↔ TiO 2 − NO 2 (4) and subsequently the NO x species undergo oxidation by the • OH radicals as follows: Photocatalysis is also used for water splitting. In this process, the electrons excited to the CB participate in the H + reduction to generate H 2 , while the holes oxidize H 2 O to form O 2. To realize the water splitting, the bottom of the CB must be more negative than the reduction potential of H + /H 2 (0 V vs. NHE (normal hydrogen electrode) at pH = 0), and the top of the VB must be more positive than the oxidation potential of H 2 O/O 2 (1.23 V vs. NHE at pH = 0). Furthermore, the band gap of the semiconductor must exceed the free energy (1.23 eV) of water splitting [48].
In another approach, photocatalysis is applied for photoreduction of CO 2 [49][50][51][52][53][54][55][56][57][58][59][60][61], resulting in the production of CO, CH 4 , CH 3 OH, HCOOH, or HCHO. The process can be realized either in the gaseous or liquid phase. In most systems, H 2 O is used as a reductant and the first step of the process is the formation of • H and H 2 . The reaction mechanism of the subsequent steps is still under debate, and the pathways depend on the process conditions and the photocatalyst properties [58].
TiO 2 photocatalysis can be also applied for the selective photocatalytic oxidation of a broad range of organic compounds, including hydrocarbons, aromatic compounds, and alcohols. The hydroxylation of aromatics is one of the examples of the photocatalytic synthesis of industrially important chemicals. Phenol (Ph), hydroquinone, and catechol are the chemicals which are widely used as precursors of resins and pharmaceutical products. Alcohols, in principle, can be photocatalytically oxidized to the corresponding carbonyl compounds [62].
Another approach is the inactivation of bacteria by the photocatalytic process. Qiu et al. [63] confirmed that the cell damage occurred and the destruction of the cytoplasmic components such as DNA and enzymes was the primary reason responsible for the bacterial inactivation.

C-Doped TiO 2
The introduction of C atoms into TiO 2 structure can lead to band gap narrowing and thus improvement in visible light absorption. The C-doped TiO 2 photocatalysts were employed for the removal of organic compounds, e.g., dyes and pharmaceuticals, from aqueous matrices [67][68][69][70], the production of hydrogen, or utilized as the bactericidal agents [71][72][73]. The development of photocatalysts containing various carbon species has recently been presented by Shi et al. [36]. The authors paid special attention to the strategies of synthesis of C-doped TiO 2 , N,C-doped TiO 2 , metal-C-doped TiO 2 , and other co-doped C/TiO 2 composites. In the present review, the C-doping of TiO 2 using different types of carbon precursors is discussed.
Qian et al. [101] presented the preparation procedure of TiO 2 -based photocatalysts with a honeycombed structure ( Figure 3). A facile template method of synthesis involves the use of PS microspheres as a template. The TiO 2 layer is formed on the surface of the template as a result of the hydrolysis process. Subsequently, calcination is applied in order to remove the template and obtain a honeycombed morphology. During the calcination, an incomplete combustion of the PS results in the formation of carbon-doped TiO 2 . There is a possibility of further processing of the as-received photocatalyst, for instance by depositing of Au onto C-doped TiO 2 layer.
Qian et al. [101] presented the preparation procedure of TiO2-based photocatalysts with a honeycombed structure (Figure 3). A facile template method of synthesis involves the use of PS microspheres as a template. The TiO2 layer is formed on the surface of the template as a result of the hydrolysis process. Subsequently, calcination is applied in order to remove the template and obtain a honeycombed morphology. During the calcination, an incomplete combustion of the PS results in the formation of carbon-doped TiO2. There is a possibility of further processing of the as-received photocatalyst, for instance by depositing of Au onto C-doped TiO2 layer. Reproduced from [101] with permission from Elsevier.
The successful incorporation of carbon atoms into the TiO2 lattice can be confirmed on the basis of X-ray photoelectron spectroscopy (XPS) analysis [76,[102][103][104]. Exemplary XPS spectra of the C-doped TiO2 are shown in Figure 4 [103]. The two strong peaks occurring usually in the range of 527-531 and 456-465 eV correspond to O 1s and Ti 2p binding energies (BE), respectively, indicating the presence of oxygen and titanium [76,[102][103][104]. The peaks centered near 530 eV are attributed either to lattice (Olat) or surface adsorbed (Osur) oxygen. The C-doping may result in a decrease in O 1s BE, thus enhancing the generation of oxygen vacancies (Ov). Qian et al. [101] reported an increase in the Olat/Osur ratio upon doping with carbon from 2.7 to 3.3. Furthermore, the introduction of carbon into the TiO2 lattice results in a simultaneous increase in Ti 3+ and Ti 4+ binding energy, observed as a shift of the Ti 2p3/2 position, which is associated with a change in the coordination environment of O and Ti atoms. The C 1s peak is observed typically at about 284-285 eV. In the example shown in Figure 4b, this peak can be further deconvoluted into two peaks corresponding to BE of 284.7 and 288.4 eV, and representing residual carbon (C-C) with sp 2 hybridization and C atoms substituting Ti in the TiO2 lattice (C-O) [76,[102][103][104]. Other peaks related to carbon species can be also present in the case of C-modified TiO2. For example, the peak assigned to the residual C=O species was observed at BE of 288.6 eV [101], while the Ti-C peak was reported at 282.3 eV [105]. Reproduced from [101] with permission from Elsevier.
The successful incorporation of carbon atoms into the TiO 2 lattice can be confirmed on the basis of X-ray photoelectron spectroscopy (XPS) analysis [76,[102][103][104]. Exemplary XPS spectra of the C-doped TiO 2 are shown in Figure 4 [103]. The two strong peaks occurring usually in the range of 527-531 and 456-465 eV correspond to O 1s and Ti 2p binding energies (BE), respectively, indicating the presence of oxygen and titanium [76,[102][103][104]. The peaks centered near 530 eV are attributed either to lattice (O lat ) or surface adsorbed (O sur ) oxygen. The C-doping may result in a decrease in O 1s BE, thus enhancing the generation of oxygen vacancies (O v ). Qian et al. [101] reported an increase in the O lat /O sur ratio upon doping with carbon from 2.7 to 3.3. Furthermore, the introduction of carbon into the TiO 2 lattice results in a simultaneous increase in Ti 3+ and Ti 4+ binding energy, observed as a shift of the Ti 2p 3/2 position, which is associated with a change in the coordination environment of O and Ti atoms. The C 1s peak is observed typically at about 284-285 eV. In the example shown in Figure 4b, this peak can be further deconvoluted into two peaks corresponding to BE of 284.7 and 288.4 eV, and representing residual carbon (C-C) with sp 2 hybridization and C atoms substituting Ti in the TiO 2 lattice (C-O) [76,[102][103][104].
Other peaks related to carbon species can be also present in the case of C-modified TiO 2 . For example, the peak assigned to the residual C=O species was observed at BE of 288.6 eV [101], while the Ti-C peak was reported at 282.3 eV [105].
an incomplete combustion of the PS results in the formation of carbon-doped TiO is a possibility of further processing of the as-received photocatalyst, for instanc positing of Au onto C-doped TiO2 layer. The successful incorporation of carbon atoms into the TiO2 lattice can be co on the basis of X-ray photoelectron spectroscopy (XPS) analysis [76,[102][103][104]. Exe XPS spectra of the C-doped TiO2 are shown in Figure 4 [103]. The two strong peak ring usually in the range of 527-531 and 456-465 eV correspond to O 1s and Ti 2p energies (BE), respectively, indicating the presence of oxygen and titanium [76,1 The peaks centered near 530 eV are attributed either to lattice (Olat) or surface ad (Osur) oxygen. The C-doping may result in a decrease in O 1s BE, thus enhancing eration of oxygen vacancies (Ov). Qian et al. [101] reported an increase in the Olat/O upon doping with carbon from 2.7 to 3.3. Furthermore, the introduction of carbon TiO2 lattice results in a simultaneous increase in Ti 3+ and Ti 4+ binding energy, obse a shift of the Ti 2p3/2 position, which is associated with a change in the coordination e ment of O and Ti atoms. The C 1s peak is observed typically at about 284-285 eV example shown in Figure 4b, this peak can be further deconvoluted into two peak sponding to BE of 284.7 and 288.4 eV, and representing residual carbon (C-C) with bridization and C atoms substituting Ti in the TiO2 lattice (C-O) [76,[102][103][104]. Othe related to carbon species can be also present in the case of C-modified TiO2. For e the peak assigned to the residual C=O species was observed at BE of 288.6 eV [101 the Ti-C peak was reported at 282.3 eV [105]. The above data reveal that carbon doping into TiO 2 can occur according to 3 various pathways: (i) substitution of O lat with C leading to a replacement of Ti-O by Ti-C bonds, (ii) replacement of Ti by C due to the rupture of Ti-O and creation of C-O bonds, or (iii) stabilization of C at the interstitial position [106,107].
Catalysts 2021, 11, 144 7 of 56 C-doped TiO 2 photocatalysts exhibit enhanced photocatalytic activity associated with changes in the crystalline structure, narrowing of the band gap, and lowering of the point of zero charge value [1]. The crystalline structure is affected by the preparation conditions of the photocatalysts, such as calcination temperature and atmosphere. In general, the presence of carbon retards the phase transformation from anatase to rutile [1]. Moreover, it was found that the crystallinity of C-doped TiO 2 nanotubes can be enhanced by annealing in argon atmosphere instead of oxygen or nitrogen [108]. Furthermore, application of Ar or N 2 was found to be advantageous in terms of improvement in photocatalytic activity. That was ascribed to the formation of O v due to C-doping, which could trap the electrons and reduce the recombination of e -/h + pairs [108]. Figure 5 presents a comparison of the pathways of electron transfer in the C-doped TiO 2 annealed under inert (Ar, N 2 ) and oxidizing (O 2 ) atmosphere. In both cases, the formation of a new state above the VB can be observed, which was attributed to the C 2p state of the interstitial carbon. Moreover, in the case of the photocatalysts annealed in the inert environment, a formation of an additional state below the CB was postulated, which was ascribed to the antibonding C-O state resulting from the creation of O v . The C-O state was not observed in the case of calcination in oxygen flow because such conditions diminished the O v [108].
The above data reveal that carbon doping into TiO2 can occur according to 3 various pathways: (i) substitution of Olat with C leading to a replacement of Ti-O by Ti-C bonds, (ii) replacement of Ti by C due to the rupture of Ti-O and creation of C-O bonds, or (iii) stabilization of C at the interstitial position [106,107].
C-doped TiO2 photocatalysts exhibit enhanced photocatalytic activity associated with changes in the crystalline structure, narrowing of the band gap, and lowering of the point of zero charge value [1]. The crystalline structure is affected by the preparation conditions of the photocatalysts, such as calcination temperature and atmosphere. In general, the presence of carbon retards the phase transformation from anatase to rutile [1]. Moreover, it was found that the crystallinity of C-doped TiO2 nanotubes can be enhanced by annealing in argon atmosphere instead of oxygen or nitrogen [108]. Furthermore, application of Ar or N2 was found to be advantageous in terms of improvement in photocatalytic activity. That was ascribed to the formation of Ov due to C-doping, which could trap the electrons and reduce the recombination of e -/h + pairs [108]. Figure 5 presents a comparison of the pathways of electron transfer in the C-doped TiO2 annealed under inert (Ar, N2) and oxidizing (O2) atmosphere. In both cases, the formation of a new state above the VB can be observed, which was attributed to the C 2p state of the interstitial carbon. Moreover, in the case of the photocatalysts annealed in the inert environment, a formation of an additional state below the CB was postulated, which was ascribed to the antibonding C-O state resulting from the creation of Ov. The C-O state was not observed in the case of calcination in oxygen flow because such conditions diminished the Ov [108]. Shao et al. [109] proposed a possible mechanism of the enhanced photocatalytic activity of C-doped TiO2 nanorods ( Figure 6). The appearance of a new impurity energy level above the VB, caused by C-doping of TiO2, results in narrowing of the band gap, and the photocatalyst can absorb the visible light. Direct promotion of electrons from the impurity level to the CB leads to the production of photogenerated electron-hole pairs. Electrons participate in the reduction of O2 molecules to form O2 • -, while holes oxidize hydroxyl ions and water molecules to • OH radicals.  The photocatalytic activity of the nonmetal-doped photocatalysts under visible light irradiation is commonly evaluated with the application of various dyes as model compounds, although such an approach was already discussed in the literature as the inappropriate method [110]. This is because dyes absorb visible light and as a result the photocatalytic reaction might be induced not only by the absorption of the visible light by the photocatalyst, but also by the photoabsorption of the radiation by the dye (i.e., dye sensitization). Nonetheless, since, in most literature reports on the visible light-active photocatalysts, the dyes are routinely applied, in this review some examples are also presented and discussed. For instance, Song et al. [100] applied the C-doped TiO2/carbon nanofibrous films (CTCNF) for photocatalytic decomposition of rhodamine B (Rh B) under visible light irradiation. After 150 min of the experiment the decolorization rate reached 66.4%-94.2%, being the lowest for the film carbonized at 900 °C and the highest for the film carbonized at 800 °C (CTCNF-800). The photocatalytic stability of the CTCNF-800 was kept constant during six cycles of recycle and reuse, with the decolorization efficiency exceeding 92%. Moreover, the authors reported good durability of the film stored for 1 year in air without illumination (90% Rh B decolorization efficiency).
The C-doped TiO2 photocatalysts were also applied for the removal of other organic compounds, such as pharmaceuticals, personal care products and even bacteria, from water [67][68][69][70]72,73], as well as for hydrogen production [71]. For instance, Shi et al. [68] prepared the carbon self-doped TiO2 flakes (CTF) with octahedral bipyramid skeleton structure and exposed {001} facet by hydrothermal treatment of TiC powder in a HF-HNO3 aqueous solution. The photocatalyst was applied for the mineralization of antibiotic ciprofloxacin (CIP) under visible light irradiation. The researchers revealed that about 70% of CIP was mineralized in 6 h of experiment. For comparison, the CIP mineralization in the presence of TiO2 P25 as a reference did not occur, while in the case of C-doped TiO2 sheets (CTS), it was about half the value (35%). According to the authors, the enhanced photoactivity of CTF was mainly caused by the synergic effect of several factors: (i) the narrowing of the band gap due to the carbon doping, (ii) the higher percentage of the exposed {001} facets, (iii) the larger specific surface area of C-doped TiO2, and (iv) its unique octahedral bipyramid skeleton structure.
The application of C-doped TiO2 for the removal of various pollutants from air was also examined [81,111]. Huang et al. [81] applied mesoporous nanocrystalline C-doped The photocatalytic activity of the nonmetal-doped photocatalysts under visible light irradiation is commonly evaluated with the application of various dyes as model compounds, although such an approach was already discussed in the literature as the inappropriate method [110]. This is because dyes absorb visible light and as a result the photocatalytic reaction might be induced not only by the absorption of the visible light by the photocatalyst, but also by the photoabsorption of the radiation by the dye (i.e., dye sensitization). Nonetheless, since, in most literature reports on the visible light-active photocatalysts, the dyes are routinely applied, in this review some examples are also presented and discussed. For instance, Song et al. [100] applied the C-doped TiO 2 /carbon nanofibrous films (CTCNF) for photocatalytic decomposition of rhodamine B (Rh B) under visible light irradiation. After 150 min of the experiment the decolorization rate reached 66.4%-94.2%, being the lowest for the film carbonized at 900 • C and the highest for the film carbonized at 800 • C (CTCNF-800). The photocatalytic stability of the CTCNF-800 was kept constant during six cycles of recycle and reuse, with the decolorization efficiency exceeding 92%. Moreover, the authors reported good durability of the film stored for 1 year in air without illumination (90% Rh B decolorization efficiency).
The C-doped TiO 2 photocatalysts were also applied for the removal of other organic compounds, such as pharmaceuticals, personal care products and even bacteria, from water [67][68][69][70]72,73], as well as for hydrogen production [71]. For instance, Shi et al. [68] prepared the carbon self-doped TiO 2 flakes (CTF) with octahedral bipyramid skeleton structure and exposed {001} facet by hydrothermal treatment of TiC powder in a HF-HNO 3 aqueous solution. The photocatalyst was applied for the mineralization of antibiotic ciprofloxacin (CIP) under visible light irradiation. The researchers revealed that about 70% of CIP was mineralized in 6 h of experiment. For comparison, the CIP mineralization in the presence of TiO 2 P25 as a reference did not occur, while in the case of C-doped TiO 2 sheets (CTS), it was about half the value (35%). According to the authors, the enhanced photoactivity of CTF was mainly caused by the synergic effect of several factors: (i) the narrowing of the band gap due to the carbon doping, (ii) the higher percentage of the exposed {001} facets, (iii) the larger specific surface area of C-doped TiO 2 , and (iv) its unique octahedral bipyramid skeleton structure.
The application of C-doped TiO 2 for the removal of various pollutants from air was also examined [81,111]. Huang et al. [81] applied mesoporous nanocrystalline C-doped TiO 2 obtained through a direct solution-phase carbonization using TiCl 4 and DEA as precursors in order to oxidize NO x in indoor air conditions under simulated solar light irradiation. The effectiveness of removal was significantly higher compared to the commercial TiO 2 P25 and amounted to 25% and 12% after 40 min of irradiation in the case of C-doped TiO 2 calcined at 500 and 600 • C, respectively.
Shim et al. [73], fabricated C-doped TiO 2 in order to inactivate Listeria monocytogenes bacteria. After 2.5 h of visible light irradiation with and without UV radiation, a 2.10 and 2.45 log reduction in bacteria was achieved, respectively. The authors stated that the photocatalytically produced ROS were responsible for the observed disinfection effect.
The C-doped TiO 2 materials were also applied for hydrogen production [71]. In the presence of the carbon-doped TiO 2 decorated with reduced graphene oxide (r-GO) and methanol as an electron donor, the H 2 production under visible light (400-690 nm) amounted to 1.50 ± 0.2 mmol g −1 h −1 [71].
In another approach, the C-doped TiO 2 was applied for photoreduction of CO 2 [59]. A series of photocatalysts were prepared by sol-gel method with the application of citric acid as a carbon source. The best performance exhibited the C-doped TiO 2 calcined at 300 • C and subsequently subjected to Al reduction treatment. The space-time yield of CH 4 and CO amounted to 4.1 and 2.5 µmol g −1 h −1 under solar light, and 0.53 and 0.63 µmol g −1 h −1 in the case of visible light irradiation, respectively.
The presented above overview shows that C-doping of TiO 2 is at the center of interest of numerous researchers. Various substrates and synthesis methods have been applied for production of C-doped TiO 2 . Among them, the most common approaches include utilization of sugars, organic acids, and (bio)polymers as carbon sources and titanium alkoxides as TiO 2 precursors, and synthesis of the photocatalysts via sol-gel, sol-microwave, and solvothermal methods. There are only a few works on less common approaches, such as those based on utilization of banana stem fibers and butterfly wings as carbon sources, or application of electrospinning for the photocatalysts fabrication. The C-doped TiO 2 was reported to have a potential application in photocatalytic processes utilizing visible or solar light. They include removal of organic contaminants from water, bacterial inactivation and hydrogen production. Nonetheless, a majority of papers are still focused on the decolorization of dyes, while degradation and mineralization of other contaminants are less extensively investigated.

N-Doped TiO 2
Among all the non-metal dopants, nitrogen is the one most frequently used [1]. This is predominantly due to its small ionization energy and its atomic size comparable with that of oxygen [112]. Numerous precursors of nitrogen have been employed for the preparation of N-doped TiO 2 , as can be seen in Table 2. Table 2. Selected examples of methods and precursors applied during preparation of the N-doped TiO 2 photocatalysts.
Makropoulou et al. [166] synthesized the N-doped TiO 2 photocatalysts using TTIP as a TiO 2 precursor and various nitrogen dopants: urea, TEA and NH 3 . A comparison of their properties is presented in Table 3. The highest specific surface area was observed for the N-doped TiO 2 prepared with urea dopant, while the lowest band gap energy value was found in the case of the photocatalyst prepared with TEA. Huang et al. [164] prepared novel N-doped anatase mesoporous bead photocatalysts using a two-cycle microwave-assisted hydrothermal method with three various nitrogen precursors: 1,6-diaminohexane (HDA), TEA and urea. They used TTIP as a TiO 2 precursor. Different nitrogen sources gave different degrees of N doping: 6.16, 1.14, and 1.24 wt.% for HDA, TEA and urea, respectively.
The as-prepared N-doped TiO 2 is typically post-annealed to improve the photocatalytic activity. Zhang et al. [185] investigated the influence of the calcination temperature on the optical and photocatalytic properties of the photocatalysts. They found that the temperature should not be lower than 200 • C. Moreover, a very high calcination temperature was reported to have an adverse effect on the TiO 2 photocatalytic activity. The highest visible-light activity of N-TiO 2 at N/Ti molar ratio of 1 was observed for the calcination temperature of 350 • C.
Krivtsov et al. [194] demonstrated that thermal treatment at temperatures of 400−500 • C caused nitrogen species to occupy interstitial positions in TiO 2 . However, the treatment under air at higher temperatures almost completely removed the N-dopants, reducing the photoactivity.
Mohamed et al. [183] examined photocatalytic activity of the N-doped TiO 2 calcined at different temperatures. They prepared the photocatalysts with the sol-gel method using HNO 3 and TBOT as N and TiO 2 precursors, respectively. The results of their research are presented in Table 4. The photocatalytic activity increased under both visible and UV irradiation as the calcination temperature increased, reaching its maximum at 400 • C. A further increase in the temperature led to a decrease in the photocatalytic activity both for visible and UV light. Moreover, the efficiency of Ph photodecomposition under visible light in the presence of the N-doped TiO 2 was visibly higher compared to that observed for commercial TiO 2 P25 and pure anatase TiO 2 (TAA). The incorporation of nitrogen resulted in a considerably narrower band gap, leading to an improved vis-light photoactivity of the modified photocatalysts compared to the undoped TiO 2 . Table 4. Properties and photocatalytic activity of the N-doped TiO 2 photocatalysts prepared at different calcination temperature, where T75 is the photocatalyst dried at 75 • C, and T200, T400, T600 and T800 are the photocatalysts calcined at 200, 400, 600, and 800 • C respectively, TAA is pure anatase TiO 2 purchased from Sigma Aldrich. Initial concentration of Ph was 50 mg L −1 [183]. There are two possible modes of nitrogen incorporation into the TiO 2 structure, interstitial (addition of nitrogen into the TiO 2 lattice) and substitutional (replacement of oxygen) [112,195]. To show the state of nitrogen, XPS analysis is performed. In the literature, the peak around 400 eV is widely assigned to the interstitial nitrogen species, directly bound to the lattice oxygen (Ti-O-N) [164,195]. According to Marques et al. [134] and Mahy et al. [165], the photoelectron lines associated with Ti-N bonds inherent to a successful substitutional doping appear in the 396-398 eV range, while lines above 400 eV are usually assigned to the interstitially doped nitrogen. Cao et al. [196] confirmed the presence of the peak at 399.4 eV, implying substitutional nitrogen doping (O-Ti-N) and the other peak at 400.8 eV, which can be assigned to the interstitial nitrogen doping (Ti-O-N) ( Figure 7). It has been reported that the interstitial N atoms give rise to the higher energy states in the band gap and behave as stronger hole trapping sites, reducing the direct oxidation ability. They are also easily oxidized in air at high temperatures [193].

Photocatalyst
sulted in a considerably narrower band gap, leading to an improved vis-light photoactivity of the modified photocatalysts compared to the undoped TiO2. Table 4. Properties and photocatalytic activity of the N-doped TiO2 photocatalysts prepared at different calcination temperature, where T75 is the photocatalyst dried at 75 °C, and T200, T400, T600 and T800 are the photocatalysts calcined at 200, 400, 600, and 800 °C respectively, TAA is pure anatase TiO2 purchased from Sigma Aldrich. Initial concentration of Ph was 50 mg L −1 [183]. There are two possible modes of nitrogen incorporation into the TiO2 structure, interstitial (addition of nitrogen into the TiO2 lattice) and substitutional (replacement of oxygen) [112,195]. To show the state of nitrogen, XPS analysis is performed. In the literature, the peak around 400 eV is widely assigned to the interstitial nitrogen species, directly bound to the lattice oxygen (Ti-O-N) [164,195]. According to Marques et al. [134] and Mahy et al. [165], the photoelectron lines associated with Ti-N bonds inherent to a successful substitutional doping appear in the 396-398 eV range, while lines above 400 eV are usually assigned to the interstitially doped nitrogen. Cao et al. [196] confirmed the presence of the peak at 399.4 eV, implying substitutional nitrogen doping (O-Ti-N) and the other peak at 400.8 eV, which can be assigned to the interstitial nitrogen doping (Ti-O-N) ( Figure 7). It has been reported that the interstitial N atoms give rise to the higher energy states in the band gap and behave as stronger hole trapping sites, reducing the direct oxidation ability. They are also easily oxidized in air at high temperatures [193].  [196] with permission from Elsevier.
The N 1s peak localized around 400 eV may also correspond to many other contributions, potentially due to impurities, usually indicating the presence of oxidized nitrogen species such as N 2 O 2 2− , NO x , N 2 , NH x or NHOH [165,172,174]. Nitrogen within nitrites (NO 2 − ) and nitrates (NO 3 − ) is commonly characterized by the N 1s with BE higher than 403 eV, while nitrogen within nitrides (N 3 − ) occurs in a range of 396-398 eV [195]. Nitrogen incorporation into TiO 2 creates an N 2p intermediate band localized above the O 2p VB. That results in a decrease in the band gap of the photocatalyst and a shift of the optical absorption to the visible light region. It was reported [39,197,198] that substitutional N-doping of TiO 2 narrows the band gap to a lesser extent than the interstitial N-doping.
The mechanism of photocatalysis in the presence of N-doped TiO 2 is summarized in Figure 8. The substitutional nitrogen state (N s ) and the interstitial nitrogen state (N i ) are directly responsible for the origin of the visible light induced photocatalysis. Nitrogen doping allows the electron to be excited from the N impurity levels (N s and N i ) to the CB of TiO 2 (paths B and C). Then the electron can be trapped by the oxygen vacancy, which is an energetically favorable process. The electrons in the CB and O v states may recombine with h + in the N impurity levels, which is an unfavorable phenomenon. Alternatively, the photoexcited e − can reduce O 2 molecules to O 2 •− , resulting further in the formation of • OH. Hydroxyl radicals can be also generated upon reaction of the holes with H 2 O or OH − . Pathway A, representing direct excitation of electrons from VB to CB of TiO 2 , is not possible under visible light irradiation [17,199].
Nitrogen incorporation into TiO2 creates an N 2p intermediate band localized above the O 2p VB. That results in a decrease in the band gap of the photocatalyst and a shift of the optical absorption to the visible light region. It was reported [39,197,198] that substitutional Ndoping of TiO2 narrows the band gap to a lesser extent than the interstitial N-doping.
The mechanism of photocatalysis in the presence of N-doped TiO2 is summarized in Figure 8. The substitutional nitrogen state (Ns) and the interstitial nitrogen state (Ni) are directly responsible for the origin of the visible light induced photocatalysis. Nitrogen doping allows the electron to be excited from the N impurity levels (Ns and Ni) to the CB of TiO2 (paths B and C). Then the electron can be trapped by the oxygen vacancy, which is an energetically favorable process. The electrons in the CB and Ov states may recombine with h + in the N impurity levels, which is an unfavorable phenomenon. Alternatively, the photoexcited ecan reduce O2 molecules to O2 •-, resulting further in the formation of • OH. Hydroxyl radicals can be also generated upon reaction of the holes with H2O or OH -. Pathway A, representing direct excitation of electrons from VB to CB of TiO2, is not possible under visible light irradiation [17,199]. A modified mechanism was proposed by Cao et al. [196] in the case of the solardriven photocatalysis over Ti 3+ and N co-doped photocatalysts. The materials were obtained by NaBH4 reduction of urea-modified mesoporous TiO2 spheres under Ar atmosphere and elevated temperature (350 °C). The mechanism is schematically illustrated in Figure 9. N-doping of TiO2 resulted in emerging of a new impurity level above the VB, which was already discussed above. Additionally, below the CB, an intermediate energy level was generated by introduction of Ti 3+ and Ov. That narrowed the band gap and consequently improved photocatalytic efficiency in a visible light. A modified mechanism was proposed by Cao et al. [196] in the case of the solar-driven photocatalysis over Ti 3+ and N co-doped photocatalysts. The materials were obtained by NaBH 4 reduction of urea-modified mesoporous TiO 2 spheres under Ar atmosphere and elevated temperature (350 • C). The mechanism is schematically illustrated in Figure 9. N-doping of TiO 2 resulted in emerging of a new impurity level above the VB, which was already discussed above. Additionally, below the CB, an intermediate energy level was generated by introduction of Ti 3+ and O v . That narrowed the band gap and consequently improved photocatalytic efficiency in a visible light.
The ratio of the substitutional to interstitial N can be changed, e.g., via the calcination procedure. Bjelajac et al. [201] reported that N-doping of TiO2 nanotubes realized by annealing in an ammonia atmosphere resulted in both substitutional and interstitial incorporation of nitrogen. However, only interstitial N was observed when a post-treatment by calcination in air was realized. It was also reported [200] that in the case of the doping of a crystalline material (strong Ti-O bonds), the substitution of oxygen with N can be . Schematic illustration for the solar-driven photocatalytic mechanism of N-doped TiO 2 (adapted from [196]).
The ratio of the substitutional to interstitial N can be changed, e.g., via the calcination procedure. Bjelajac et al. [201] reported that N-doping of TiO 2 nanotubes realized by annealing in an ammonia atmosphere resulted in both substitutional and interstitial incorporation of nitrogen. However, only interstitial N was observed when a post-treatment by calcination in air was realized. It was also reported [200] that in the case of the doping of a crystalline material (strong Ti-O bonds), the substitution of oxygen with N can be prevented, which results in a lower concentration of substitutional N. On the opposite, the amorphous structure of TiO 2 (weak Ti-O bonds) facilitates substitutional N-doping.
Nonetheless, the role of the interstitial and substitutional nitrogen in the improvement in the photocatalytic activity of N-doped TiO 2 is still unclear. Some authors reported that interstitial N exhibited better photoactivity under visible irradiation than substitutional N [202,203], while others reported that the substitutional N improves the photocatalytic activity to a higher extent [204,205]. Moreover, there are works showing that the visible photocatalytic activity is better when the quantity of the interstitial and substitutional N is equal [200].
Monteiro et al. [209] prepared a series of visible light-active N-doped TiO 2 photocatalysts by a simple impregnation method, well suited for a scale-up to mass production. They used urea and TiO 2 P25 as precursors. Photocatalytic activity of the obtained photocatalysts was tested by degradation of diphenhydramine hydrochloride (DP) under visible light irradiation and was found to depend on the amount of nitrogen dopant and calcination temperature. The optimal calcination temperature was found to be 380 • C. The photocatalyst prepared at the urea/TiO 2 ratio of 0.5 exhibited the highest photocatalytic efficiency ( Figure 10). This was explained by the lowest band gap energy obtained for that photocatalyst (2.99 eV).
Xing et al. [174] used a common antibiotic CIP to evaluate the photocatalytic properties of the N-doped TiO 2 immobilized on the glass spheres. The photocatalytic activity experiments were carried out with 20 mL CIP aqueous solution (20 mg L −1 ) and the supported photocatalyst loading was equal to 3 g L −1 . A 90% removal rate in 90 min under visible light irradiation from a 500 W xenon lamp (λ > 420 nm) was reported. Moreover, the removal rate of CIP in the five cycles of reuse was more than 90% in 120 min and almost no exfoliated photocatalyst was observed.
Delavari et al. [139] synthesized the N-doped titania nanotube arrays via electrochemical anodization method. The photocatalyst was applied to photocatalytic conversion of CO 2 and CH 4 . With optimal experimental conditions of 250 W UV light power, 2 cm distance between UV lamp and reactor and initial amount of 10% CO 2 , the efficiency of photoconversion was up to 41.5% and 62.2% for CO 2 and CH 4 , respectively.
The N-doped TiO 2 was also used in the photocatalytic production of hydrogen [168,222]. Reddy et al. [168] prepared various N-doped TiO 2 photocatalysts using the hydrothermal method followed by calcination in air at 350 • C. Among them, the one with multiphase titania consisting of 69% anatase, 17% brookite and 14% rutile showed the highest photoactivity towards hydrogen production (10.5 mmol h −1 g −1 ). Preethi et al. [222]  Xing et al. [174] used a common antibiotic CIP to evaluate the photocatalytic properties of the N-doped TiO2 immobilized on the glass spheres. The photocatalytic activity experiments were carried out with 20 mL CIP aqueous solution (20 mg L −1 ) and the supported photocatalyst loading was equal to 3 g L −1 . A 90% removal rate in 90 min under visible light irradiation from a 500 W xenon lamp (λ > 420 nm) was reported. Moreover, the removal rate of CIP in the five cycles of reuse was more than 90% in 120 min and almost no exfoliated photocatalyst was observed.
Delavari et al. [139] synthesized the N-doped titania nanotube arrays via electrochemical anodization method. The photocatalyst was applied to photocatalytic conversion of CO2 and CH4. With optimal experimental conditions of 250 W UV light power, 2 cm distance between UV lamp and reactor and initial amount of 10% CO2, the efficiency of photoconversion was up to 41.5% and 62.2% for CO2 and CH4, respectively.
The N-doped TiO2 was also used in the photocatalytic production of hydrogen [168,222]. Reddy et al. [168] prepared various N-doped TiO2 photocatalysts using the hydrothermal method followed by calcination in air at 350 °C. Among them, the one with multiphase titania consisting of 69% anatase, 17% brookite and 14% rutile showed the highest photoactivity towards hydrogen production (10.5 mmol h −1 g −1 ). Preethi et al. [222]  The N-doped TiO 2 microsheets fabricated hydrothermally from TiN in the presence of HF and HCl were successfully applied for CO 2 photoreduction [60]. After 2 h of UV-vis or visible light irradiation, the production rate of CH 3 OH reached 0.71 µmol g −1 and 0.28 µmol g −1 , respectively, and was significantly higher than observed for TiO 2 P25 (0.05 µmol g −1 and 0 µmol g −1 , respectively).
The above overview confirms that the N-doped TiO 2 is the most widely investigated photocatalyst among all of the non-metal-doped titania with enhanced visible light photocatalytic activity. The incorporation of nitrogen into the TiO 2 structure is relatively easy due to its atomic size, comparable with that of oxygen. Various methods of modification of TiO 2 with nitrogen have been applied, including both the wet (hydrolysis, sol-gel, solvothermal, plasma enhanced electrolysis) and dry (magnetron sputtering, CVD, PLD) approach. The most common nitrogen sources are urea, ammonia, nitric acid, N 2 gas and TEA, although numerous other compounds were also applied. Nitrogen-doped TiO 2 exhibits a broad absorption in the visible region, which allows the utilization of a large part of the solar spectrum. This is caused by a reduction in the band gap resulting from the interstitial or substitutional incorporation of nitrogen in TiO 2 . The N-doped TiO 2 was proposed to be applied for effective water and air purification, as well as energy production, or medical and antimicrobial applications.
Moreover, newer synthetic strategies have also been proposed. Pillai et al. [267] used continuous ball milling with moderate thermoannealing (Figure 11), which resulted in a unique S-doped TiO 2 nanostructure composed of macroporous channels with mesoporous cores. This environmentally friendly, solventless and template-free method requires titanium hydroxide and elemental sulfur as substrates.
Moreover, newer synthetic strategies have also been proposed. Pillai et al. [267] used continuous ball milling with moderate thermoannealing (Figure 11), which resulted in a unique S-doped TiO2 nanostructure composed of macroporous channels with mesoporous cores. This environmentally friendly, solventless and template-free method requires titanium hydroxide and elemental sulfur as substrates. Figure 11. Schematic representation of the procedure of preparation of macroporous S-doped TiO2 photocatalysts. Reprinted with permission from [267]. Copyright 2020 American Chemical Society.
Sharotri et al. [249] investigated the influence of sulfur doping on the phase transition of TiO2. In order to study that, they prepared the S-doped TiO2 photocatalysts calcined at 750 °C with different molar ratios of EDT (S source) to TiO2 (1:0.1, 1:0.5 and 1:1). The results (Table 6) confirm that sulfur doping leads to a delay in the phase transition from anatase to rutile.

EDT/TiO2 Molar Ratio
Phase Composition Figure 11. Schematic representation of the procedure of preparation of macroporous S-doped TiO 2 photocatalysts. Reprinted with permission from [267]. Copyright 2020 American Chemical Society.
Sharotri et al. [249] investigated the influence of sulfur doping on the phase transition of TiO 2 . In order to study that, they prepared the S-doped TiO 2 photocatalysts calcined at 750 • C with different molar ratios of EDT (S source) to TiO 2 (1:0.1, 1:0.5 and 1:1). The results ( Table 6) confirm that sulfur doping leads to a delay in the phase transition from anatase to rutile. There are two possible routes of sulfur doping-cationic or anionic. It has been reported that cationic doping takes place by titanium substitution with S 4+ or S 6+ , while anionic doping occurs by oxygen substitution with S 2− [272]. The chemical state of the doped sulfur can be investigated by the XPS analysis ( Figure 12). Generally, a peak at the BE values above 168 eV in the S 2p spectra can be assigned to sulfur on the higher oxidation state (S 4+ , S 6+ ), substituting Ti atoms in the lattice of TiO 2 [236,260]. The peaks around 160-163 eV can indicate the Ti-S bonds' formation due to the substitution of O atoms in the TiO 2 lattice by S 2− [236,246,251,273].
There are two possible routes of sulfur doping-cationic or anionic. It has been reported that cationic doping takes place by titanium substitution with S 4+ or S 6+ , while anionic doping occurs by oxygen substitution with S 2− [272]. The chemical state of the doped sulfur can be investigated by the XPS analysis ( Figure 12). Generally, a peak at the BE values above 168 eV in the S 2p spectra can be assigned to sulfur on the higher oxidation state (S 4+ , S 6+ ), substituting Ti atoms in the lattice of TiO2 [236,260]. The peaks around 160-163 eV can indicate the Ti-S bonds' formation due to the substitution of O atoms in the TiO2 lattice by S 2− [236,246,251,273]. The presence of cationic sulfur in the S-doped TiO2 photocatalysts was confirmed in the literature by the peaks positioned at 168.3 [246], 168.5 [232], or 168.9 eV [233]. Zhu et al. [248] claimed that the peak at 168.5 eV, corresponding to S 6+ , can be further deconvoluted into two peaks at 168.5 and 169.7 eV (Figure 13), which can be attributed to S 2p3/2 and S 2p1/2, respectively, with S 2p3/2 peak about twice the intensity or area higher than the S 2p1/2 peak. Doping S 6+ into TiO2 structure can form an impurity energy level above the VB, resulting in a reduction in the TiO2 band gap [260].
Anionic sulfur doping is more difficult to achieve since the bond strength of the existing Ti-O bond (672.4 KJ/mol) is larger than Ti-S bond (418.0 KJ/mol). Additionally, the ionic radius of S 2− (1.7 Å) has a higher value than O 2− (1.22 Å) [246]. As a result, the incorporation of sulfur by substitution of Ti 4+ with S 6+ is energetically more favorable than the replacement of O 2− by S 2− [248]. The presence of cationic sulfur in the S-doped TiO 2 photocatalysts was confirmed in the literature by the peaks positioned at 168.3 [246], 168.5 [232], or 168.9 eV [233]. Zhu et al. [248] claimed that the peak at 168.5 eV, corresponding to S 6+ , can be further deconvoluted into two peaks at 168.5 and 169.7 eV (Figure 13), which can be attributed to S 2p 3/2 and S 2p 1/2 , respectively, with S 2p 3/2 peak about twice the intensity or area higher than the S 2p 1/2 peak. Doping S 6+ into TiO 2 structure can form an impurity energy level above the VB, resulting in a reduction in the TiO 2 band gap [260]. Bakar et al. [236] compared photodecomposition efficiency of methyl orange (MO) over 2ATO and 2CTO S-doped TiO2. The photocatalysts were prepared via template free and low-temperature OPM followed by crystallization through hydrothermal treatment. Thiourea was reported to be mainly responsible for cationic S-doping due to the substitution of Ti 4+ by S 6+ , while CS2 promoted the anionic doping via substitution of O 2− by S 2− . Both, cationic and anionic S-doped photocatalysts revealed the enhanced visible light photocatalytic activity. However, S 6+ was found to be responsible for reducing the TiO2 crystallite size and in the cationic S-doped TiO2 the photoinduced holes and chemisorbed hydroxyls played a major role during photocatalysis. On the other hand, the S 2− doping resulted in an increased TiO2 crystallite size, and in anionic S-doped TiO2, the photoinduced holes and electrons played a nearly equal role in the photocatalytic activity. Bakar et al. [236] compared photodecomposition efficiency of methyl orange (MO) over 2ATO and 2CTO S-doped TiO 2 . The photocatalysts were prepared via template free and low-temperature OPM followed by crystallization through hydrothermal treatment. Thiourea was reported to be mainly responsible for cationic S-doping due to the substitution of Ti 4+ by S 6+ , while CS 2 promoted the anionic doping via substitution of O 2− by S 2− . Both, cationic and anionic S-doped photocatalysts revealed the enhanced visible light photocatalytic activity. However, S 6+ was found to be responsible for reducing the TiO 2 crystallite size and in the cationic S-doped TiO 2 the photoinduced holes and chemisorbed hydroxyls played a major role during photocatalysis. On the other hand, the S 2− doping resulted in an increased TiO 2 crystallite size, and in anionic S-doped TiO 2 , the photoinduced holes and electrons played a nearly equal role in the photocatalytic activity.
The mechanism of photocatalysis in the presence of the cationic and anionic S-doped TiO 2 was proposed by various researchers [236,275]. Ma et al. [275] obtained the cationic doping by the application of thiourea as sulfur precursor, while the usage of CS 2 resulted in both cationic and anionic doping ( Figure 14).
substituting oxygen, (ii) interstitial sulfur (SO)O, and (iii) a combination of (S)O and (SO)O in which interstitial sulfur shares a lattice site with (S)O forming (S2)O. The authors reported that under conditions poor in oxygen, the substitutional sulfur (S)O is dominant, while under conditions rich in oxygen, the interstitial sulfur (SO)O becomes more stable. In other words, the incorporation of (S)O may be easier than (SO)O under oxygen-poor conditions, which is due to a lower formation energy.
Bakar et al. [236] compared photodecomposition efficiency of methyl orange (MO) over 2ATO and 2CTO S-doped TiO2. The photocatalysts were prepared via template free and low-temperature OPM followed by crystallization through hydrothermal treatment. Thiourea was reported to be mainly responsible for cationic S-doping due to the substitution of Ti 4+ by S 6+ , while CS2 promoted the anionic doping via substitution of O 2− by S 2− . Both, cationic and anionic S-doped photocatalysts revealed the enhanced visible light photocatalytic activity. However, S 6+ was found to be responsible for reducing the TiO2 crystallite size and in the cationic S-doped TiO2 the photoinduced holes and chemisorbed hydroxyls played a major role during photocatalysis. On the other hand, the S 2− doping resulted in an increased TiO2 crystallite size, and in anionic S-doped TiO2, the photoinduced holes and electrons played a nearly equal role in the photocatalytic activity.
The mechanism of photocatalysis in the presence of the cationic and anionic S-doped TiO2 was proposed by various researchers [236,275]. Ma et al. [275] obtained the cationic doping by the application of thiourea as sulfur precursor, while the usage of CS2 resulted in both cationic and anionic doping ( Figure 14).  The improvement of the photocatalytic activity of the cationic S-doped TiO 2 was attributed to [261,275]: (i) the formation of a midgap level under the CB due to the incorporation of S 6+ , leading to the band gap narrowing and increase in the visible-light absorption; (ii) the enhanced adsorption of OH − on the photocatalyst surface resulting from the created charge imbalance in the bulk of TiO 2 upon replacement of Ti 4+ by S 6+ . These adsorbed hydroxyl ions and the OH − ions present in the bulk solution can be oxidized by h + to form highly oxidative • OH radicals; and (iii) the formation of Ti-O-S bonds, which due to the higher electronegativity of O atoms leads to a partial transfer of electrons from sulfur to oxygen. As a result, the electron-deficient S atoms can capture the e − , which reduces the recombination of electron-hole pairs and improves the quantum efficiency. In the case of the cationic and anionic S-doped TiO 2 , the enhancement of the photoactivity was explained by: (i) the formation of Ti-O-S and O-Ti-S bands in the crystal lattice under the cationic and anionic S-doping, respectively, which introduce new impurity levels between the VB and CB. As a result, the formation of electron-hole pairs under the visible light radiation is improved; (ii) the generation of O v and defects in the TiO 2 lattice, which contribute to the improvement of photoactivity by (a) enhancement of visible light absorption and (b) capture of electrons thus inhibiting the e − -h + recombination; (iii) the increased generation of hydroxyl and superoxide radicals by the reaction of h + and e − with the adsorbed H 2 O and O 2 , respectively [261,275]. Moreover, an important role of the surface-adsorbed SO 4 2− groups in charge separation due to acting as electron traps was also proposed [261].
The photocatalytic activity of S-doped TiO 2 under visible light is usually investigated using dyes as model compounds. Table 7 presents the properties of different thiourea S-doped TiO 2 prepared using various titania precursors and the efficiency of methylene blue (MB) decomposition under visible light in the presence of these photocatalysts. Wang et al. [256] examined the effect of synthesis temperature on the visible light photocatalytic activity of the S-doped TiO 2 prepared by hydrothermal oxidation of TiS 2 (ST-120-4) or its high temperature calcination in air (ST-550-10). Rh B was used as a model compound. As can be seen from the plot (Figure 15), about 50% of Rh B was decomposed using TS-550-10 and 80% using TS-120-4 within 60 min. This difference in the photocatalytic activity was attributed to the significantly larger surface area of ST-120-4, afforded by the low-temperature hydrothermal synthesis. The surface area of ST-550-10 was 6.6 m 2 g −1 , while that of ST-120-4 was 46.4 m 2 g −1 , i.e., seven times greater. eration of hydroxyl and superoxide radicals by the reaction of h + and ewith the adsorbed H2O and O2, respectively [261,275]. Moreover, an important role of the surface-adsorbed SO4 2− groups in charge separation due to acting as electron traps was also proposed [261].
The photocatalytic activity of S-doped TiO2 under visible light is usually investigated using dyes as model compounds. Table 7 presents the properties of different thiourea Sdoped TiO2 prepared using various titania precursors and the efficiency of methylene blue (MB) decomposition under visible light in the presence of these photocatalysts.
Wang et al. [256] examined the effect of synthesis temperature on the visible light photocatalytic activity of the S-doped TiO2 prepared by hydrothermal oxidation of TiS2 (ST-120-4) or its high temperature calcination in air (ST-550-10). Rh B was used as a model compound. As can be seen from the plot (Figure 15), about 50% of Rh B was decomposed using TS-550-10 and 80% using TS-120-4 within 60 min. This difference in the photocatalytic activity was attributed to the significantly larger surface area of ST-120-4, afforded by the low-temperature hydrothermal synthesis. The surface area of ST-550-10 was 6.6 m 2 g −1 , while that of ST-120-4 was 46.4 m 2 g −1 , i.e., seven times greater.  Boningari et al. [261] investigated the photodegradation of acetaldehyde over a cationic S-doped TiO 2 in the presence of visible light. In order to obtain the photocatalysts, sulfuric acid at various concentrations (1-3 mol L −1 ) was applied as S source, while TTIP was used as TiO 2 precursor. The presence of both S 6+ /S 4+ oxidation states, corresponding to the BE of 167.5-167.9 and 169.1-169.6 eV, respectively, was noted. The highest content of S calculated on the basis of XPS was observed for the photocatalyst prepared using 2 mol L −1 H 2 SO 4 (8.9 at%). For this photocatalyst, the S 6+ /S 4+ ratio was also the highest (5.61). Nevertheless, the most efficient total organic carbon (TOC) removal (ca. 60%) was found in the case of the photocatalyst fabricated from 3 mol L −1 H 2 SO 4 (containing 5.19 at% of S).
Bakar et al. [236] applied the anionic and cationic S-doped TiO 2 for the decomposition of colored MO and colorless Ph under visible light irradiation. The rate of MO decolorization after 240 min of the experiment reached 89.32% for 2ATO and 85.43% for 2CTO. A significant mineralization efficiency was also found for both photocatalysts. Moreover, the authors reported that the efficiency of Ph photodegradation and mineralization was similar to that observed for MO. It was also noted that the pristine TiO 2 exhibited no significant photoactivity under visible light.
Sraw et al. [268] prepared S-doped TiO 2 photocatalyst via wet impregnation method. The authors demonstrated the degradation process of two persistent and highly toxic organophosphorus pesticides, monocrotophos (MCP) and quinalphos (QP). The S-doped photocatalyst showed better performance under the sunlight than under UV irradiation. The removal efficiency of MCP (25 mg L −1 ) was 95.36% and 68.21% for sunlight and UV, respectively. In the case of QP (20 mg L −1 ), a similar effectiveness of decomposition was observed (98.09% and 68.67%, respectively). The increase in photocatalytic activity under the sunlight was due to the changes in the band gap energy of the photocatalyst caused by doping with sulfur. Examination of different S:Ti ratios (Table 8) revealed that the optimum value was 0.7:1, in the case of which the band gap energy was lowered by about 10% compared to the undoped TiO 2 . Moreover, the photocatalyst with the S:Ti ratio of 0.7:1 had the lowest average particle size (27.2 nm), as was determined by the transmission electron microscopy (TEM). Table 8. Effect of S-dopant concentration on the E bg and average particle size of S-doped TiO 2 [268]. In another case [269], a photodegradation of atrazine was investigated. The applied Sdoped TiO 2 photocatalysts were obtained from TiCl 4 and (NH 4 ) 2 SO 4 , and contained sulfur in S 6+ state. The studies revealed higher effectiveness of atrazine oxidation by the solar/Sdoped TiO 2 system in comparison with the solar/TiO 2 system. After 30 min of irradiation, the removal rate reached 60% and 40%, respectively. The authors also emphasized that lower water pH favored the photodegradation of herbicide, while the presence of humic acids (HA) decreased the decomposition rate due to the competition with the atrazine molecules for hydroxyl radicals and partial absorption of sunlight before it reached TiO 2 .

Photocatalyst E bg (eV) Average Particle Size by TEM (nm)
In turn, Lin et al. [271] examined the S-doped TiO 2 synthesized from TTIP and thiourea, in terms of degradation of gaseous 1,2-dichloroethane. The S-doped TiO 2 photocatalysts exhibited superior photocatalytic activity under visible light compared to that of pure TiO 2 . The conversion rate of 1,2-dichloroethane was 55.3 nmol min −1 g −1 , whereas in the case of the undoped TiO 2 it was 2.16 nmol min −1 g −1 after 1 h of irradiation. Among all the examined photocatalysts, the highest efficiency of photodecomposition was found for the S-doped TiO 2 containing 2.46 wt.% of sulfur.
The S-doped TiO 2 synthesized by Baeissa [230] was applied for photooxidation of cyanide in water (KCN concentration: 25-200 mg L −1 ). The photocatalytic activity of the S-doped TiO 2 increased with increasing sulfur content from 0 to 0.3 wt.%. In the case of the highest S content, the efficiency of photocatalytic oxidation of cyanide reached 100% after 30 min of irradiation with a blue fluorescent lamp (150 W). The stability of the photocatalyst was confirmed during five cycles of recycling and reuse.
Yi et al. [234] proposed the sol-gel method followed by calcination at 300 • C to prepare the S-doped TiO 2 for visible light photocatalytic degradation of diclofenac (DCF) (10 mg L −1 ). The efficiency of DCF decomposition reached 93% after 4 h for 0.8 g L −1 of photocatalyst. A significant role of h + , • OH, and O 2 •− in the DCF degradation was confirmed in the experiments involving various scavengers: (a) benzoquinone (BQ) as a superoxide radical scavenger, (b) AgNO 3 as electron capturer, (c) tert-butyl alcohol as an • OH scavenger, and (d) methanol (MeOH) and potassium iodide (KI) as both • OH and h + scavengers.
Furthermore, the S-doped TiO 2 showed potential for application in antimicrobial field. Dunnill et al. [279] investigated the inhibition of E. coli growth in the presence of the S-doped TiO 2 films. The samples were irradiated with the white light source, typically used in the hospitals. After 24 h of irradiation, a 99.5% inhibition of bacterial growth (i.e., 2.3 log 10 reduction) was observed.
Other researchers [61] described a significant improvement in the photoreduction of CO 2 using S-doped TiO 2 under UV-A and visible light irradiation. A series of photocatalysts was obtained via simple sonothermal method using sulfur powder as a dopant source. For the most active S-doped TiO 2 , the evolution of 6.25 µmol g −1 of methane, 2.74 µmol g −1 of ethylene, 0.074 µmol g −1 of propylene and 0.030 µmol g −1 of propane after 24 h of UV-A irradiation in KOH solution was reported. Moreover, generation of methane and methanol in the acetonitrile-water mixture was observed under visible light, with the yield of 167.6 and 12,828.4 µmol g −1 , respectively.
Based on the analysis of the literature data, it can be concluded that doping of TiO 2 with sulfur is not as common as doping it with nitrogen, but still is one of the major strategies of non-metal modification of TiO 2 aimed at improvement of its visible light photoactivity. The main source of sulfur is thiourea, although many other compounds have also been used, such as DMSO, TiOSO 4 , and sulfur powder. A majority of the photocatalysts were prepared with the use of titanium alkoxides as TiO 2 precursors. Among numerous synthesis methods, sol-gel and hydrothermal treatment are the most common approaches. The two possible routes of sulfur doping are cationic (S 6+ /S 4+ ) or anionic (S 2− ), with the former being more energetically favorable and thus more often reported. It was found that S-doped TiO 2 materials exhibit superior properties as antibacterial agents. Moreover, this type of photocatalyst was employed in the decolorization of dyes and decomposition of both organic and inorganic contaminants.

C,N-Co-Doped TiO 2
The C,N-co-doped TiO 2 can be obtained by application of various carbon and nitrogen sources. A summary of the precursors is shown in Table 9. Among all nitrogen precursors, ammonia was the most commonly used, while in the case of the carbon source, various researchers applied different compounds ( Table 9). The most popular Ti sources were TTIP and TBOT. The modified photocatalysts were obtained by numerous approaches, including sol-gel, solvothermal and hydrothermal methods, calcination, pyrolysis, or even green bioinspired synthesis.
According to Ananpattarachai et al. [89] the C,N-co-doped TiO 2 combines the synergistic effect of the C-doped and N-doped TiO 2 . Due to the carbon doping, the pollutant could be easier adsorbed on the surface of TiO 2 . The carbonaceous specie I formed by the incorporated C atoms could act as a photosensitizer (C*). After the excitation of this specie, electrons can be injected into the CB of TiO 2 . Further, these electrons could be transferred to O 2 adsorbed on the photocatalyst surface, leading to the formation of the superoxide anion radicals (O 2 •− ), which could be subsequently transformed to H 2 O 2 and • OH, and together with the oxidative species formed by the oxidation of H 2 O and OH − by holes participate in pollutants degradation ( Figure 16). Nitrogen doping could lead to the narrowing of the band gap by creation of an intra band gap states (IB) above the VB, enabling the visible light absorption. Furthermore, nitrogen doping can result in a shift of the flat-band (FB) potential position to a higher level than in the case of an unmodified TiO 2 .  Ti foil CBD MO (5 mg L −1 ) 100% in 30 min, visible light [306] electrons can be injected into the CB of TiO2. Further, these electrons could be transferred to O2 adsorbed on the photocatalyst surface, leading to the formation of the superoxide anion radicals (O2 •-), which could be subsequently transformed to H2O2 and • OH, and together with the oxidative species formed by the oxidation of H2O and OHby holes participate in pollutants degradation ( Figure 16). Nitrogen doping could lead to the narrowing of the band gap by creation of an intra band gap states (IB) above the VB, enabling the visible light absorption. Furthermore, nitrogen doping can result in a shift of the flatband (FB) potential position to a higher level than in the case of an unmodified TiO2. To investigate the state of N and C in the modified photocatalysts the XPS analysis is typically conducted. Table 10   To investigate the state of N and C in the modified photocatalysts the XPS analysis is typically conducted. Table 10 presents a summary of the nitrogen and carbon species detected in the XPS spectra of the C,N-co-doped TiO 2 . Usually, the N 1s signals in the range of ca. 396-402 eV are attributed to the presence of O-Ti-N bonds, indicating the substitutional incorporation of N atom into the TiO 2 lattice [89,280,298,300,302,303,[305][306][307][308]. The presence of carbon is confirmed by the peaks ranging from 284.6 to 289.6 eV ( Table 10). The most commonly reported signals are the peaks at 285.9-286.7 and 288.5-288.7 eV, which are mainly ascribed to C-O and C=O bonds [89,280,289,296,298,302,303,305,306], and the signals at 288.0-288.9 eV attributed to the Ti-O-C bonds [89,288,296,297,301,303,306]. Table 10. An overview of N 1s and C 1s signals reported in the XPS spectra of the C,N-co-doped TiO 2 .
Xiao et al. [284] applied the C-modified and N-doped TiO 2 hollow spheres for the degradation of tetracycline (TC) and TCH under simulated solar irradiation. The degradation rate constants k reached 0.1812 and 0.1785 min −1 , respectively. Moreover, the authors reported that after four cycles of reusing, the material exhibited almost unchanged TC degradation efficiency.
Recently, the pollution of the environment with microplastics has become a matter of concern. Ariza-Tarazona et al. [287] prepared C,N-co-doped TiO 2 photocatalysts and investigated the impact of pH and temperature on the degradation process of the highdensity polyethylene (HDPE) microplastics under visible light. The influence of these factors is shown in Figure 17. The results prove that low pH and low temperature had a combined positive effect on the microplastics' degradation. This could be due to the presence of H + ions facilitating the degradation process. Low temperature contributed to the microplastics' fragmentation, increasing their surface area and interaction with the photocatalyst. The authors obtained over 70% average mass loss at pH 3 and temperature of 0 • C. density polyethylene (HDPE) microplastics under visible light. The influence of these factors is shown in Figure 17. The results prove that low pH and low temperature had a combined positive effect on the microplastics' degradation. This could be due to the presence of H + ions facilitating the degradation process. Low temperature contributed to the microplastics' fragmentation, increasing their surface area and interaction with the photocatalyst. The authors obtained over 70% average mass loss at pH 3 and temperature of 0 °C. Another proposed application is photocatalytic H2 production. Zhang et al. [308] prepared the C,N-co-doped TiO2 hollow spheres using TiCN as a precursor of anatase TiO2 and a self-doping source of C and N. The photocatalyst was synthesized by a one-pot hydrothermal method. The resulting C,N-co-doped TiO2 showed enhanced photocatalytic activity towards H2 production under the visible light irradiation. The H2 generation rate was 445 times higher than for pure TiO2.
The modification of TiO2 with nitrogen and carbon was also found to be an efficient approach to enhance its antibacterial activity [282]. The photocatalysts were prepared using gaseous NH3 as a nitrogen source and aliphatic alcohols (methanol, ethanol and isopropanol) as carbon sources. The C,N-co-doped TiO2 was incorporated into concrete plates. The plates were irradiated with an artificial solar light in the presence of Escherichia coli bacteria. A complete inactivation of bacteria after 60 min of irradiation was observed Another proposed application is photocatalytic H 2 production. Zhang et al. [308] prepared the C,N-co-doped TiO 2 hollow spheres using TiCN as a precursor of anatase TiO 2 and a self-doping source of C and N. The photocatalyst was synthesized by a one-pot hydrothermal method. The resulting C,N-co-doped TiO 2 showed enhanced photocatalytic activity towards H 2 production under the visible light irradiation. The H 2 generation rate was 445 times higher than for pure TiO 2 .
The modification of TiO 2 with nitrogen and carbon was also found to be an efficient approach to enhance its antibacterial activity [282]. The photocatalysts were prepared using gaseous NH 3 as a nitrogen source and aliphatic alcohols (methanol, ethanol and isopropanol) as carbon sources. The C,N-co-doped TiO 2 was incorporated into concrete plates. The plates were irradiated with an artificial solar light in the presence of Escherichia coli bacteria. A complete inactivation of bacteria after 60 min of irradiation was observed in the case of the photocatalysts treated at 100 • C in the presence of ethanol and isopropanol and treated at 300 • C in the presence of methanol. The authors reported that not only inactivation but also damage of the bacterial cells was achieved.
The review of the current state of the art revealed that the simultaneous doping with carbon and nitrogen is a predominately used mode of co-doping of TiO 2 with C, N and S non-metals. A majority of the photocatalysts were synthesized on the base of titanium alkoxides as TiO 2 precursors, similarly as in the case of C-and N-doped TiO 2 . Numerous methods have been applied; some of them were common, such as sol-gel, solvothermal or calcination under gas flow, but there were also others such as pyrolysis, electrospinning, and CBD. Various alcohols are usually used as carbon sources. The most common nitrogen source is ammonia, although many other compounds were used, for example compounds containing amine groups such as glycine, HMT, or polyaniline. Moreover, some researchers used a single compound as both C and N source. Among others, PAN, TELA and urea were used for this purpose. The C,N-co-doped TiO 2 has been successfully applied to decompose various organic pollutants and hydrogen production. It was also used as an antibacterial agent.

C,S-Co-Doped TiO 2
The literature reports on the C,S-co-doped TiO 2 are very limited. There are only a few papers on that subject and a majority of them were published in the last 5 years. The summary of substrates and methods applied for the preparation of the C,S-co-doped TiO 2 is shown in Table 11. Rate constants: up to 4.04 × 10 4 s −1 for UV and 0.63 × 10 4 s −1 for vis [157] Thiourea, urea TBOT Hydrolysis 4-chlorophenol (1.5 × 10 −4 mol L −1~3 5% in 3 h, visible light [312] Ivanov et al. [156], Khalyavka et al. [157], and Romanovska et al. [310] used the widely applied thiourea as a dopant of C and S. Although the authors [310] claimed their photocatalyst to be C,S-co-doped TiO 2 , the XPS analysis revealed the signals characteristic for C, N and S (285, 400, and 170 eV, respectively). The calculated surface content of the non-metals was 16.7 at% for C, 0.4 at% for S and 0.3 at% for N. Nonetheless, since the elemental analysis showed no presence of nitrogen, the authors attributed the observed weak XPS signal in the N 1s region to the adsorption of gaseous N 2 and NH 3 from the atmosphere.
The entirely new approaches were proposed by El Nemr et al. [102] and Chaudhary et al. [309]. The method reported by El Nemr et al. [102] was an ecofriendly procedure based on the sol-gel synthesis under alkaline conditions in the presence of Eichhornia crassipes aqueous leaf extract, followed by calcination in air at 400 • C. Except for deionized water, no other solvents were applied. Chaudhary et al. [309] proposed the application of the sulfate rich seaweed polysaccharides, carrageenans, (kappa (κ-), iota (ι-) and lambda (λ-)) from red seaweeds as a source of sulfur and carbon, while TTIP was used as TiO 2 substrate. Sulfur was present in two oxidation states, S 4+ and S 0 , corresponding to the XPS BE of 168 and 164.4 eV, respectively. It was observed that application of calcination at 600 • C led to a conversion of S 0 to S 4+ , as was concluded from a disappearance of the 164.4 eV peak and the presence of an intense peak at~168 eV. The authors concluded that no new Ti-S bond was formed at the expense of Ti-O bonds, and that the sulfate ions could form O-S-O bonds on the TiO 2 surface. The presence of C 1s peak at 287.4 eV was also observed and it was assigned to the graphitic carbon.
Ivanov et al. [156] proposed a more conventional approach in which metatitanic acid was applied as a TiO 2 precursor, while thiourea was the source of carbon and sulfur. A homogenous mixture of the substrates was calcined at 500 • C. The obtained C,S-co-doped TiO 2 contained sulfur exclusively in the oxidation state of +6 (XPS BE of 168.8 eV). In the case of carbon, three peaks at BE of 285, 287 and 289 eV were observed. The signal at 285 eV was attributed to an external post-synthesis (not in vacuo) contamination. The other two peaks were assigned to C-O and C=O bonds indicating the presence of carbon substituting lattice Ti atoms and forming a Ti-O-C structure or C=O groups adsorbed on the photocatalyst surface. The surface content of sulfur and carbon was visibly higher (2.8 and 30.3 at%, respectively) than the content in the powder volume (0.36 and 5.07 at%, respectively).
Photocatalytic activity of the C,S-co-doped TiO 2 was analyzed mainly on the basis of the removal of various dyes, including Rh B [156,313], RB19 [102], RR76 [102], MB [156] and ST [157]. El Nemr et al. [102] investigated the photoactivity of the C,S-co-doped TiO 2 with reference to RB19 and RR76 decolorization in a real textile wastewater and the toxicity of the treated and untreated effluents. A 400 W halide lamp was used as a radiation source. The results were compared with those obtained in the presence of C-doped TiO 2 and S-doped TiO 2 . It was reported that although all the doped photocatalysts exhibited higher photocatalytic activity than pure TiO 2 , the efficiency of decolorization was higher for the C-doped TiO 2 than for the other photocatalysts. After 30 min of irradiation the concentration of RB19 decreased by 5.5%, 77.8%, 72.2%, and 73.3%, while the concentration of RR76 was lowered by 7.5%, 86.7%, 73.3%, and 63.3%, in the presence of TiO 2 , C-TiO 2 , C,S-TiO 2 and S-TiO 2 , respectively. However, the analysis of chemical oxygen demand revealed that the highest degradation efficiency was obtained in the presence of C,S-codoped TiO 2 (97%), while in the case of the C-doped TiO 2 and S-doped TiO 2 , it amounted to 91%. The toxicity of the treated wastewater samples analyzed with reference to rotifer and Artemia salina was less than 4%, being significantly lower compared to that of the untreated wastewater (95-98%). On the basis of the tests conducted in the presence of scavengers, it was concluded that the main species participating in the decomposition of the dyes were photogenerated holes, not the hydroxyl radicals. Moreover, the obtained photocatalyst was successfully recovered and reused at least three times without significant decrease in photoactivity.
In another case [157], the dye ST was photocatalytically degraded under UV and visible (568 nm) light in the presence of a series of C,S-co-doped TiO 2 with various dopants content. The decolorization rates were in the range of 0.68-4.04 × 10 4 s −1 for UV and 0.20-0.63 × 10 4 s −1 for visible light, depending on the photocatalyst composition. In the case of the pure TiO 2 , the decolorization rate was 2.58 × 10 4 s −1 for UV, while no visible decomposition was observed for vis. The authors concluded that the following effects were responsible for the enhanced dye removal: (i) interfacial bonding, (ii) defective sites, (iii) narrowing of the band gap, (iv) inhibition of electron-hole pair recombination in the presence of carbon and sulfur, (v) prolongation of the lifetime of the photogenerated charges, (vi) formation of doping electronic states and (vii) change in textural characteristics.
Hohol et al. [313] applied the C,S-co-doped TiO 2 as well as TiO 2 P25 as the modifiers of the cement mortars. The photocatalytic activity was analyzed on the basis of Rh B removal in the visible light. The sample containing 2 wt.% of the C,S-co-doped TiO 2 revealed the highest photocatalytic activity corresponding to the 87% removal efficiency after 2 h of irradiation. The cement mortars modified with 1 wt.% of the C,S-co-doped TiO 2 , 2 wt.% of TiO 2 P25 and 1 wt.% of TiO 2 P25 showed a removal efficiency of 65, 44 and 37%, respectively.
The overview presented above reveals that the number of papers on C,S-co-doped TiO 2 is very limited. However, it is worth noting again here the case of thiourea precursor. As was already explained, this compound is used for the preparation of the S-doped TiO 2 , C,S-co-doped TiO 2 , or even C,N,S-tri-doped TiO 2 , thus in some cases it is difficult to distinguish between these types of single-and multiple doping. Thus, further investigations on the C,S-co-doped TiO 2 are necessary. The studies should be especially related to the synthesis method and type of precursors, the structural and electronic properties of the photocatalysts and their effect on the photocatalytic activity analyzed with reference to other compounds than dyes.

N,S-Co-Doped TiO 2
The simultaneous doping with nitrogen and sulfur is considered as an efficient method to obtain a reduction in the band gap of TiO 2 [314]. Thiourea [231,247,276,[315][316][317][318][319][320] as an S source and urea [87,231,315,316,318] or ammonium nitrate [276] as an N source are usually proposed. There are also reports on the application of ammonium sulfate [321] and ammonium thiocyanate [322] as both N and S precursors. Albrbar et al. [247] applied an aprotic solvent DMSO as a precursor for S-doping, while the annealing in ammonia stream was applied for N-doping.
Some authors performed a comparative evaluation of the non-metal doped TiO 2 . The N,S-co-doped TiO 2 was compared with the N-doped and S-doped photocatalysts. A summary of these data is shown in Table 12. Sol-gel and wet impregnation were mainly applied as synthesis methods. In general, co-doping resulted in a smaller crystallite size of anatase and lower band gap energy values than observed in the case of the singledoped photocatalysts. The effect of the co-doping on the S BET was dependent on the synthesis method. An increase in the specific surface area compared to the single co-doped photocatalysts was observed only in the case of the sol-gel approach.  Figure 18 shows the effects of the N and S dopants on the visible light photocatalytic activity of the N,S-co-doped TiO 2 photocatalyst. According to Chung et al. [314], nitrogen forms a delocalized state in the band gap of TiO 2 leading to the enhanced solution bulk reaction by increasing the visible light absorption. On the other hand, sulfur promotes the adsorption of organic species present in the solution onto the surface of the N,S-co-doped TiO 2 , thereby enhancing the surface reaction with cationic organic pollutants. Figure 18 shows the effects of the N and S dopants on the visible light photocatalytic activity of the N,S-co-doped TiO2 photocatalyst. According to Chung et al. [314], nitrogen forms a delocalized state in the band gap of TiO2 leading to the enhanced solution bulk reaction by increasing the visible light absorption. On the other hand, sulfur promotes the adsorption of organic species present in the solution onto the surface of the N,S-co-doped TiO2, thereby enhancing the surface reaction with cationic organic pollutants. The XPS analysis of the N,S-co-doped TiO2 synthesized by the non-hydrolytic sol-gel approach using DMSO and gaseous NH3 as S and N precursors, respectively, revealed interstitial nitrogen doping and cationic sulfur doping by the substitution of Ti 4+ with S 4+ and S 6+ [247]. The N 1s peak was observed at ~400 eV. The authors noticed that the interstitial doping of N 3-ion into TiO2 lattice is typical for annealing of titania in ammonia atmosphere. The S 2 p signal was in the form of a doublet with a maximum of the S 2p3/2 at 169.6 eV. It was also noted that annealing in ammonia suppresses the oxidation of sulfur, and as a result the substitution of Ti 4+ with S 4+ and S 6+ is less favorable compared to calcination in air [247]. The presence of interstitial N and substitution of Ti 4+ with S 4+ /S 6+ Figure 18. Schematic illustration of the suggested effects of the nitrogen and sulfur dopants on the enhanced visible light photocatalytic activity of the N,S-co-doped TiO 2 (adapted from [314,[323][324][325][326]).
The XPS analysis of the N,S-co-doped TiO 2 synthesized by the non-hydrolytic sol-gel approach using DMSO and gaseous NH 3 as S and N precursors, respectively, revealed interstitial nitrogen doping and cationic sulfur doping by the substitution of Ti 4+ with S 4+ and S 6+ [247]. The N 1s peak was observed at~400 eV. The authors noticed that the interstitial doping of N 3ion into TiO 2 lattice is typical for annealing of titania in ammonia atmosphere. The S 2 p signal was in the form of a doublet with a maximum of the S 2p 3/2 at 169.6 eV. It was also noted that annealing in ammonia suppresses the oxidation of sulfur, and as a result the substitution of Ti 4+ with S 4+ and S 6+ is less favorable compared to calcination in air [247]. The presence of interstitial N and substitution of Ti 4+ with S 4+ /S 6+ was also reported in the case of application of ammonium thiocyanate as a dopant [322]. In contrast, Abu Bakar et al. [316] reported the absence of the interstitial nitrogen and the presence of the anionic N-doping in the case of N,S-co-doped TiO 2 obtained by the OPM-assisted hydrothermal procedure using ammonia, thiourea and urea as N and S precursors. The substitution of O atoms by N atoms in the TiO 2 lattice led to the formation of O-Ti-N bonds. The main form of sulfur was S 6+ substituting Ti 4+ , as was found on the basis of the XPS peak at 168.7 eV. Moreover, a weak signal at the BE of 163.8 eV was observed indicating that anionic doping due to the replacement of oxygen by S 2− in the crystal lattice of TiO 2 also occurred. The above data clearly show that the synthesis route strongly affects the type of doping.
The photocatalytic activity of the N,S-co-doped TiO 2 was examined using dyes such as Rh B [314,316,324], MB [291], MO [317,322,326], reactive orange 16 (RO16) [247], as well as HA [87,231], Ph [276,316], TC [325], non-steroidal anti-inflammatory drugs such as ibuprofen (IBP) and naproxen (NPX) [323], anti-cancer drug 5-fluorouracil (5-FU) [327], or an organophosphorus flame retardant tris(1-chloro-2-propyl)phosphate (TCPP) [321]. Table 13 presents selected examples of the application of the N,S-co-doped TiO 2 photocatalysts in organic pollutants removal. Eslami et al. [323] prepared the N,S-co-doped TiO 2 NPs and nanosheets through facile sol-gel and hydrothermal methods, respectively. They used thiourea as nitrogen and sulfur source, and TBOT as the TiO 2 precursor. The photocatalytic activity of the two types of the modified TiO 2 was examined in IBP and NPX decomposition under visible light ( Figure 19). The NPs displayed better photocatalytic activity by degrading 85% of IBP and 99.3% of NPX, whereas 71.6% of IBP and 99.1% of NPX were degraded by nanosheet structures. Both photocatalysts exhibited higher degradation and mineralization efficiency than TiO 2 P25. The N,S-co-doped TiO2 nanowires (NWs) were synthesized by Zhang et al. [318] via hydrothermal approach using titanium sulfate, thiourea and urea as the precursors. The experiments exhibited that the N,S-co-doped NWs can be easily separated from aqueous suspension by a simple sedimentation within ca. 30 min, while TiO2 P25 suspension remains turbid at this time. Furthermore, the N,S-co-doped NWs were proved to maintain their photocatalytic activity during five cycles of atrazine degradation under visible light. The authors concluded that the ease of separation and the stable photocatalytic performance make the prepared photocatalyst an attractive candidate for the industrial water purification.
The influence of the climate conditions on the photocatalysis process with N,S-co- The N,S-co-doped TiO 2 nanowires (NWs) were synthesized by Zhang et al. [318] via hydrothermal approach using titanium sulfate, thiourea and urea as the precursors. The experiments exhibited that the N,S-co-doped NWs can be easily separated from aqueous suspension by a simple sedimentation within ca. 30 min, while TiO 2 P25 suspension remains turbid at this time. Furthermore, the N,S-co-doped NWs were proved to maintain their photocatalytic activity during five cycles of atrazine degradation under visible light. The authors concluded that the ease of separation and the stable photocatalytic performance make the prepared photocatalyst an attractive candidate for the industrial water purification.
The influence of the climate conditions on the photocatalysis process with N,S-codoped TiO 2 was studied by Khalilian et al. [326]. Thiourea was applied as a source of nitrogen and sulfur while TBOT was used as TiO 2 precursor. The authors compared the degradation efficiency of MO under natural sunlight irradiation in cloudy and clear sky. The MO decomposition rate was 96% after 3 h in clear sky conditions, while in the case of the cloudy sky it reached 90% after 4.5 h. A more significant difference was observed after a shorter irradiation time, e.g., after 1 h, the decolorization rate amounted to about 75% and 30%, respectively.
Ju et al. [317] applied hydrothermal treatment followed by calcination for the preparation of the N,S-co-doped TiO 2 photocatalysts from Ti(SO 4 ) 2 and thiourea as precursors. Various N/Ti atomic ratio and calcination temperature were used. The highest photocatalytic activity during MO removal under sunlight exhibited the photocatalyst obtained at 3% of N and calcined at 500 • C. The authors attributed the highest photoactivity of this photocatalyst to the small crystallite size of anatase (7 nm). The analysis of the influence of the pH of MO solution on its decolorization rate revealed the highest removal efficiency (92% after 6 h) at pH 4.
Yan et al. [324] prepared the N,S-co-doped TiO 2 films by S-doping with thiourea of the previously N-doped TiO 2 nanograss array films on Ti wire mesh obtained through moderate chemical oxidation. The films were applied for Rh B decolorization under visible light irradiation. The decomposition rate constants calculated for the undoped, N-doped and N,S co-doped TiO 2 amounted to 9.77 × 10 −4 , 3.39 × 10 −3 and 1.70 × 10 −2 , respectively. These data show that the photocatalytic activity of the pure TiO 2 was increased by 2.5 times due to N-doping, while the subsequent S-doping enhanced it by 4.0 times. The N,S-co-doped TiO 2 film kept its photocatalytic activity under visible light during five cycles of reuse, while in the case of the N-doped TiO 2 , a gradual deterioration of the photodecolorization rate was observed.
Birben et al. [231] compared the photocatalytic performance of the N-doped TiO 2 , Sdoped TiO 2 and S,N-co-doped TiO 2 synthesized according to the wet impregnation method with the use of TiO 2 P25 (about 80% anatase and 20% rutile) or Hombikat UV-100 (100% anatase) as TiO 2 sources. The photocatalytic activity of the prepared photocatalysts was assessed regarding the photocatalytic degradation of HA in terms of decolorization and mineralization. The experiment was conducted in the reactor equipped with a 250 W m −2 Xenon lamp (wavelength range of 300-800 nm). Initial concentrations of HA and photocatalysts were 20 mg L −1 and 0.25 g L −1 , respectively. The results reveal that the S,N-co-doped TiO 2 prepared with the use of Hombikat UV-100 had the highest photocatalytic degradation efficiency among all the examined photocatalysts. The pseudo-first order rate constants k calculated for color removal and mineralization amounted to 11.4 × 10 −2 min −1 and 6.80 × 10 −2 min −1 , respectively. In the case of the S,N-co-doped TiO 2 P25, the photocatalytic performance was significantly worse (k = 2.28 × 10 −2 min −1 and 2.81 × 10 −2 min −1 , respectively).
Another application of the N,S-co-doped TiO 2 photocatalyst was described by Antonopoulou et al. [321]. The authors prepared the photocatalysts via the sol-gel method using TBOT as a TiO 2 precursor and ammonium sulfate as a source of S and N. Different N,S:Ti molar ratios of 0.5, 1, and 1.5 were applied. The photocatalytic efficiency of the obtained materials was evaluated during photodegradation of an organophosphorus flame retardant TCPP under simulated solar light (Xenon lamp (2.2 kW) with 290 nm cut-off glass filter) and under visible light (>400 nm, 20 W led flood lamp emitting 1600 lumens). The highest photodegradation efficiency under visible light was observed in the case of the N,S-co-doped TiO 2 obtained with the equimolar N,S:Ti ratio. The pseudo-first order rate constant k amounted to 0.64 × 10 −3 min −1 . For the simulated solar light, however, the photodegradation was more efficient in the presence of the single N-doped TiO 2 than in the case of the N,S-co-doped TiO 2 with the rate constants of 2.6 × 10 −3 and 1.6 × 10 −3 min −1 , respectively. Nonetheless, the lowest photodegradation of TCPP was observed for the undoped TiO 2 (0.01 × 10 −3 and 0.31 × 10 −3 min −1 for visible and simulated solar light, respectively).
The N,S-co-doped TiO 2 was also proposed to be applied in lithium storage. Jiao et al. [328] compared the lithium storage capacity of the undoped, N-doped, S-doped and N,S-co-doped TiO 2 as active anode materials. The N-doped TiO 2 revealed only slightly (3.5 times) higher storage capacity than the undoped TiO 2 , while the S-doped TiO 2 was characterized by even lower efficiency than TiO 2 . However, co-doping of TiO 2 with N and S led to a significantly (17 times) higher storage capacity compared to the undoped TiO 2 . The superior performance of the N,S-co-doped TiO 2 compared to the undoped and single doped TiO 2 was ascribed to its significantly improved electronic conductivity resulting from a modified electronic structure of the semiconductor due to a homogenous distribution of the N and S dopants.
The presented above overview shows that co-doping of TiO 2 with N and S can enhance the photocatalytic activity of titania under visible light due to the formation of new impurity levels. The N,S-co-doped TiO 2 photocatalysts were prepared mainly via sol-gel and wet impregnation approaches. The major reported source of sulfur is thiourea, while the applied nitrogen sources include primarily ammonia, ammonium nitrate and urea. It remains unclear what the role of thiourea is as a source of nitrogen in this context. Furthermore, the authors did not discuss the presence of carbon, which is another element in the thiourea structure. Hence, an additional research on this subject is essential. The N,S-co-doped TiO 2 photocatalysts were successfully employed in the photodecomposition of organic compounds, mostly dyes, but also pharmaceuticals and HA.

C,N,S-Tri-Doped TiO 2
Single doping of TiO 2 with carbon, nitrogen, or sulfur results in an improved photocatalytic activity under visible light. Tri-doping of TiO 2 has been proposed as a more efficient method of enhancement of photoactivity and extending the visible light absorption range of titania photocatalysts [7,8,[329][330][331][332][333][334][335][336][337]. Sushma et al., in their review [39], summarized the progress on C,N,S-tri-doped TiO 2 in the years 2008-2015. They compared the characteristics of the photocatalysts regarding preparation method and further temperature treatment, the type of dopant as well as TiO 2 precursor, the band gap energy obtained from UV-vis/DR spectra and the dopant states acquired from XPS. The most common methods were sol-gel and hydrothermal treatment. TBOT and TTIP were often used as the TiO 2 precursors, while thiourea and cystine were the most frequently applied sources of C, N and S dopants. In Table 14, there is a continuation of the overview covering the articles from the year 2015 to the present.
Catalysts 2021, 11, x FOR PEER REVIEW 37 of 58 as a photosensitizer, injecting an electron into the CB (compare Figure 16), whereas incorporation of nitrogen and sulfur leads to mixing of the O 2p orbitals of TiO2 with N 2p and S 2p orbitals. As a result, the band gap is narrower compared to the undoped TiO2, which allows the direct electron excitation into CB using the visible light. Additionally, the anionic superoxide radials O2 •could be generated by a reduction of the adsorbed O2 molecules with CB electrons, and afterwards hydroxyl radicals • OH can be formed. According to the authors [7], in the case of the C,N,S-tri-doped TiO2, this route of hydroxyl radicals formation plays a significant role.  Another mechanism was proposed by Amreetha et al. [334] (Figure 21) in the case of a colorful compound-an organic dye Rh B. They suggested three pathways of the photodecolorization. The first pathway refers to the dye photosensitization process and is based on the excitation of the dye to the triplet excited state under the visible light irradiation. Furthermore, the dye molecules are converted to cationic dye radicals and the electrons are injected to the TiO 2 CB according to Equations (8)-(10): Catalysts 2021, 11, x FOR PEER REVIEW 37 of 58 as a photosensitizer, injecting an electron into the CB (compare Figure 16), whereas incorporation of nitrogen and sulfur leads to mixing of the O 2p orbitals of TiO2 with N 2p and S 2p orbitals. As a result, the band gap is narrower compared to the undoped TiO2, which allows the direct electron excitation into CB using the visible light. Additionally, the anionic superoxide radials O2 •could be generated by a reduction of the adsorbed O2 molecules with CB electrons, and afterwards hydroxyl radicals • OH can be formed. According to the authors [7], in the case of the C,N,S-tri-doped TiO2, this route of hydroxyl radicals formation plays a significant role. Another mechanism was proposed by Amreetha et al. [334] (Figure 21) in the case of a colorful compound-an organic dye Rh B. They suggested three pathways of the photodecolorization. The first pathway refers to the dye photosensitization process and is based on the excitation of the dye to the triplet excited state under the visible light irradiation. Furthermore, the dye molecules are converted to cationic dye radicals and the electrons are injected to the TiO2 CB according to Equations (8)-(10): dye + hν → dye * (8) dye * + TiO → dye  + TiO e O + e CB → O  (10) Figure 21. Electron migration in C,N,S-tri-doped TiO2 with binary polymorphs (adapted from [334]).
Electrons contribute to the production of the anionic superoxide radicals O 2 •− , followed by the formation of hydroperoxyl radicals • OOH and then H 2 O 2 , which further dissociates into hydroxyl radicals • OH. The hydroxyl radicals participate in the decomposition and mineralization of the dye. The second pathway is similar to the mechanism proposed by Ramandi et al. [7]. However, Amreetha et al. [334] discussed it with reference to a mixed phase material composed of anatase and rutile. The electrons from the intermediate states created due to N and S doping (N 2p, S 2p) are excited to the CB of TiO 2 . Due to the presence of the mixed phases the charge separation occurs which limits the negative phenomenon of electron-hole recombination. There are two possible methods of electron transfer: (i) from the CB of anatase to the CB of rutile and (ii) from the CB edge of rutile to the lattice trapping sites of anatase localized below the CB of both anatase and rutile. At the same time, the hole transfer from the VB of anatase to the VB of rutile occurs. The third pathway assumes carbon to behave like a photosensitizer, which injects an electron into the CB (compare Figure 20). Table 15 presents a brief summary of photocatalytic applications of the C,N,S-tridoped TiO 2 . It was used to decompose dyes such as Rh B [330,333], MB [337,340], MO [335] and Rose Bengal [329]. Another common application was the removal of pharmaceutical compounds such as DCF [7], IBP [8], and TC [338] from water. Chun et al. [332] used the C,N,S-tri-doped TiO 2 for the decomposition of volatile organic compounds (VOCs) such as benzene, toluene, ethyl benzene and o-xylene. Tri-doped TiO 2 was also applied for decomposition of NO x [339,342]. Khedr et al. [8] revealed about 40 times higher initial rate of IBP photodegradation under visible light with application of the C,N,S-tri-doped TiO 2 (1.779 µmol min −1 ) in comparison with the undoped TiO 2 (0.043 µmol min −1 ). The efficiency of mineralization was 95.2% and 2.8% for the tri-doped and unmodified photocatalysts, respectively.
Wang et al. [338] prepared several C,N,S-tri-doped TiO 2 photocatalysts via sol-gel method using different thiourea to Ti molar ratio (0.03:1, 0.05:1, 0.10:1 and 0.15:1) and various calcination temperature (350, 450, 550 and 650 • C). The authors tested the photocatalytic activity of the obtained photocatalysts in the TC removal process under visible light irradiation. The highest photocatalytic activity (97% removal after 3 h) exhibited the photocatalyst calcined at 450 • C with the thiourea to Ti molar ratio of 0.05:1. The surface content of the non-metals in that photocatalyst was 12.56 at% of C, 1.60 at% of S, and 0.54 at% of N, as was found from the XPS analysis. For comparison, the commercial TiO 2 P25 achieved 26% removal efficiency. The authors attributed the superior performance of the C,N,S-tri-doped TiO 2 to the synergistic effects of (i) the high adsorption of TC associated with the high specific surface area of the photocatalyst, (ii) the narrowing of the band gap resulting from the C,N,S tri-doping, (iii) the presence of carbonaceous species which acted as a photosensitizer, and (iii) the good crystallinity of the anatase phase.
The C,N,S-tri-doped single crystal black TiO 2 nanosheets with exposed {001} facets synthesized by the hydrothermal in situ solid state reduction approach were applied for the photocatalytic decolorization of MO and evolution of hydrogen under visible light. After 180 min of irradiation, the MO decomposition rate reached 92.13% and was significantly higher than that obtained in the presence of the undoped TiO 2 (42.38%). Similarly, the highest H 2 evolution rate (149.7 µmol h −1 g −1 ) was observed for the C,N,S-tri-doped single crystal black TiO 2 . The authors asserted the high stability of the photocatalyst, confirmed by the constant hydrogen evolution rate during five cycles of reuse [335].
Bengtsson et al. [339] applied the C,N,S-tri-doped TiO 2 synthesized from TTIP and thiourea for the degradation of NO x in air under visible light. The influence of the calcination temperature (400-650 • C) and thiourea content on the properties and activity of the photocatalysts was especially investigated. It was found that the increase in the calcination temperature improved the photoactivity under visible light, while in the case of UV irradiation, the correlation was not clear. Moreover, it was reported that a key role in the decomposition of NO x under visible light was played by the presence of nitrogen doped in the TiO 2 structure. On the contrary, an increase in sulfur content resulted in a decrease in the efficiency of NO x degradation. The effect of the carbon amount was ambiguous and it was postulated than an optimal content exists (16-20 at.%), above and below which the presence of carbon has a negative influence on the photoactivity. Moreover, no correlation between the band gap and the photocatalytic activity under visible light was observed.
The data presented above show that in recent years, only a few groups of researchers worked on the C,N,S-tri-doped TiO 2 photocatalysts. Most of them used thiourea as the C, N and S source. Thiourea was also utilized for the preparation of S-doped, C,S-co-doped and N,S-co-doped TiO 2 . Thus, it remains vague as to what is the role of this precursor in the synthesis of single-and co-doped photocatalysts. The C,N,S-tri-doped TiO 2 was obtained mainly through sol-gel and hydrothermal procedures, although wet impregnation and sonochemical approaches were also proposed. Regardless of the synthesis method, sulfur was typically incorporated in the TiO 2 crystal lattice in the S 6+ oxidation state, as was concluded on the basis of the XPS analysis. Incorporation of carbon was usually confirmed by the peaks related to the presence of C-O and C=O species. In the case of nitrogen, both doping modes, i.e., substitutional or interstitial, were reported, depending on the preparation procedure. The C,N,S-tri-doped TiO 2 materials have been proved to exhibit a superior photocatalytic performance towards the photodecomposition of organic compounds such dyes and pharmaceuticals in water, as well as the photodegradation of VOCs or NO x in air.

Conclusions and Perspectives
The present review is focused on the progress of the research on the doping, co-doping and tri-doping of TiO 2 with C, N, or S. It confirms the great potential of this approach for the fabrication of visible light-active photocatalysts. The influence of the methods of the photocatalysts synthesis, including the applied precursors and process parameters, on the physicochemical properties and photoactivity of the doped TiO 2 is discussed. Moreover, the mechanisms of photocatalysis in the presence of the single-, co-or tri-doped TiO 2 are presented. Numerous applications of the non-metal doped TiO 2 are also summarized, including the removal of organic compounds from water/wastewater, treatment of air, production of hydrogen, lithium storage, or inactivation of bacteria.
The most widespread approaches of the synthesis of the C, N or S-doped TiO 2 photocatalysts are sol-gel, hydrothermal, solvothermal and wet impregnation methods, while the major precursors of TiO 2 , carbon, nitrogen and sulfur are titanium alkoxides, sugars, urea, and thiourea, respectively. However, great ambiguity in the case of the application of thiourea as a modifying agent exists. Thiourea has been used as the source of sulfur in S-doped, C,S-co-doped, and N,S-co-doped TiO 2 or as the source of carbon and sulfur in the C,S-co-doped TiO 2 , while for the C,N,S-tri-doped TiO 2 it was the precursor of all three non-metals. Thus, it remains vague as to what the role of thiourea is and which of the three mentioned above non-metals are built in the structure of the modified TiO 2 photocatalysts. This leads to the conclusion that further thorough research regarding the effect of not only thiourea, but also other C, N and S sources on the structure and properties of the doped TiO 2 is essential.
Numerous studies have shown that the incorporation of non-metals into TiO 2 usually results in the narrowing of the band gap due to the formation of new impurity levels (C 2p, N 2p, S 2p) above the VB of the semiconductor. As a result, a red shift of optical absorption leading to an enhancement of the visible light photocatalytic activity is commonly reported. Moreover, the presence of C, N, or S could also contribute to the increase in the specific surface area or the improvement in crystallinity, thus additionally enhancing the photocatalytic performance.
The most common method of doping of TiO 2 is modification with nitrogen. This issue has been widely investigated over the years. Presently, there is growing interest in the doping of TiO 2 with more than one non-metal, including C,N-, C,S-, and N,S-co-doping or C,N,S-tri-doping. Two possible modes of nitrogen incorporation can occur, depending on the preparation conditions, i.e., interstitial (Ti-O-N) and substitutional (O-Ti-N) doping. In the case of sulfur, anionic (as S 2− ) or cationic (as S 4+ or S 6+ ) doping is possible, with the latter case being more energetically favorable and, thus, more commonly reported. The modification with carbon includes the widest range of species that could possibly be formed, such as Ti-C, C-C, C-N, C=N, C=O or C-O, etc. The reported pathways of Cdoping include: (i) substitution of lattice oxygen with carbon (formation of Ti-C bonds); (ii) replacement of Ti by C (formation of C-O bonds); or (iii) stabilization of C at the interstitial position. The various mechanisms of doping affect the properties of the photocatalysts, although clear correlations between the modification procedure and the type of doping difficult to find. Therefore, more extensive investigations regarding this issue are necessary, especially when the photocatalysts with designed properties are considered.
It is not possible to unambiguously indicate the most advantageous mode of non-metal doping of TiO 2 . Each of them has some advantages and disadvantages. Moreover, a clear correlation between the doping mode and the physicochemical properties or photoactivity of the modified TiO 2 is difficult to find. Co-doping and tri-doping of TiO 2 can result in combining the properties of the particular single doped TiO 2 , leading to the enhancement of photoactivity. However, such improvement usually requires the employment of more reagents and more complicated synthesis procedures. Moreover, only a few authors compared the co-doped or tri-doped photocatalysts with single doped ones, while most of the others referred their results to an undoped TiO 2 . On the other hand, comparing the photocatalysts obtained by different authors would not be reliable since various conditions of experiments were applied, e.g., light sources and intensity, type and concentration of model pollutant, or dose of photocatalysts. Therefore, simple and reliable standard methods for testing of photocatalytic activity that will be widely applied by scientists around the world are a key issue to enable the comparison of the results obtained in different laboratories. That would contribute to defining the correlations and development of methods for designing the highly active photocatalysts.
A majority of applications of doped photocatalysts refer to the removal of pollutants from water and wastewater. The C-, N-, and S-doped photocatalysts were applied mainly for the decomposition of various dyes. Considering that the main aim of modification of TiO 2 with the non-metals is the enhancement of its visible light photoactivity, such an attempt is not appropriate, as was already widely discussed in the literature. This is because of the dye sensitization effect. Therefore, it is very important to study the photocatalytic activity with the application of colorless compounds, such as phenols, pharmaceuticals, etc. Moreover, the determination of mineralization efficiency apart from decomposition rates is also important. A detailed evaluation of the mechanisms of the visible light photocatalysis, with reference to the by-product formation or the role of various ROS is also of high significance. Finally, the toxicity of the treated solutions should be carefully examined as one of the most important parameters reflecting the treatment efficiency and environmental safety. The important research should also focus on applications of doped photocatalysts other than water/wastewater treatment. The visible light photocatalytic air treatment, hydrogen production, CO 2 photoreduction, and bacterial inactivation are still not thoroughly examined.
The enhanced photocatalytic activity under visible light irradiation makes the doped TiO 2 a promising material for low-cost photocatalytic processes. Nonetheless, although the modified photocatalysts show superior properties compared to the undoped TiO 2 , more investigations are still needed to develop the production methods at the industrial scale. The crucial matter is to obtain stable and reusable photocatalysts using simple and economically feasible procedures.

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