Titanium dioxide (TiO2) nanomaterials are widely considered to be state-of-the-art photocatalysts for environmental protection and energy conversion. However, the low photocatalytic efficiency caused by large bandgap and rapid recombination of photo-excited electrons and holes is a challenging issue that needs to be settled for their practical applications. Structure engineering has been demonstrated to be a highly promising approach to engineer the optical and electronic properties of the existing materials or even endow them with unexpected properties. Surface structure engineering has witnessed the breakthrough in increasing the photocatalytic efficiency of TiO2 nanomaterials by creating a defect-rich or amorphous surface layer with black color and extension of optical absorption to the whole visible spectrum, along with markedly enhanced photocatalytic activities. In this review, the recent progress in the development of black TiO2 nanomaterials is reviewed to gain a better understanding of the structure-property relationship with the consideration of preparation methods and to project new insights into the future development of black TiO2 nanomaterials in photocatalytic applications.

1. Introduction

Titanium dioxide (TiO2) nanomaterials have been considered as the most promising semiconductor photocatalysts for pollutant removal and energy generation owing to their relatively good photocatalytic activity, low cost, nontoxicity, and high stability since the discovery of hydrogen evolution through the photoelectrochemical water splitting on TiO2 electrode [13]. Meanwhile, photocatalysis potentially can be an ideal solution to current environmental issues and energy crisis by only consuming solar energy. Over the past decades, nanotechnology has greatly contributed to the development of TiO2 materials in photocatalysis across the globe [27]. However, the large bandgaps (3.0–3.2 eV over different phases) of TiO2 nanomaterials limit their optical absorption to ultraviolet (UV) light, along with the rapid recombination of photo-excited electrons and holes, resulting in low photocatalytic efficiency [111]. Extending the utilization of solar energy to visible region has thus been the urgent need for practical applications of TiO2 nanomaterials.

The optical and electronic properties of solid materials highly depend on the structure including the way the atoms are bonded and arranged, the phases and their distribution, and the defects [1216]. Therefore, tuning these states in solid materials can potentially tailor the optical and electronic properties of the existing solid materials [1719]. Since the discovery of visible-light active nitrogen-doped TiO2 [19], structural modification of TiO2 nanomaterials has been at the research frontier to extend the utilization of solar light to visible region, while other methods, such as localized surface plasmon resonance of plasmonic nanostructures [18, 20, 21], have been exploited. In 2011, the electronic band structure of black TiO2 nanocrystals was reported to be largely narrowed for massive visible light absorption and conversion to chemical energy, leading to markedly enhanced photocatalytic activity towards photocatalytic pollution removal and hydrogen generation from water [22]. Since then black TiO2 nanomaterials have attracted unprecedented interest in visible light utilization. Over the past decade, many methods, such as hydrogenation, aluminum reduction, and chemical reduction, have been developed to synthesize black TiO2 nanomaterials. In this review, we focus on the recent research progress in black TiO2 nanomaterials for photocatalytic applications, especially for photocatalytic hydrogen generation and pollutant removal. The basic properties of black TiO2 nanomaterials are first discussed in brief, and then, typical examples are given for each preparation method. We aim to get a better understanding on the relationship between the structure and photocatalytic properties of black TiO2 nanomaterials from different preparation methods, along with considering the effect of preparation methods on their structures. Finally, summary and perspectives on the development of black TiO2 nanomaterials are addressed.

2. Basic Properties of Black TiO2 Versus White TiO2

TiO2 exists in different phases. Anatase, rutile, brookite, and TiO2(B) are the four main polymorphs of TiO2. It is widely acknowledged that rutile is thermodynamically the most stable phase, while anatase, brookite, and TiO2(B) are metastable phases. The metastable phases of TiO2 will go through phase transformation at elevated temperatures. Usually, it is considered that anatase goes through phase transformation to rutile [23, 24] and brookite and TiO2(B) to anatase then to rutile [25, 26]. However, exceptions have been reported. Direct transformation of brookite to rutile is observed, while in the case of the anatase–brookite mixture, anatase transforms firstly to brookite and then to rutile [27]. It is suggested that the phase transformation process is affected by various parameters.

Generally, anatase and rutile TiO2 nanomaterials can be easily prepared using conventional preparation methods, while the preparation of brookite and TiO2(B) nanomaterials needs special attention. Anatase is relatively stable, and the phase transformation between anatase and rutile is dependent on a variety of factors, such as particle size and defects in anatase TiO2 [28, 29]. So far, anatase and rutile TiO2 nanomaterials are the most widely studied photocatalysts.

Different phases have different structures. Generally, all four types of polymorphs mentioned above comprise of TiO6 octahedra, but differ in the distortion of the octahedron units and share edges and corners in different manners (Figure 1) [25]. The differences in TiO6 octahedra arrangement result in different physicochemical properties and thus different photocatalytic activities. To the best of our knowledge, black TiO2 nanomaterials have been prepared from white anatase, rutile, brookite, and TiO2(B). It should be noted that phase transformation was observed in a very few cases when white TiO2 nanomaterials converted into black ones during the modification process [23].

2.1. Basic Structure

Color change from white to black for TiO2 nanomaterials reflects the change in optical properties and thus suggests the structural change after modification, at least the surface structural change. High-resolution transmission electron microscope (HRTEM) has revealed that a number of black TiO2 nanomaterials have a crystalline/amorphous core/shell structure, while white ones have clear lattice fringes throughout the crystals [22, 23, 3044]. The amorphous or disordered surface layer has been considered the typical feature of black TiO2 nanomaterials. Figure 2 shows the HRTEM images of black and white TiO2 nanomaterials with Figure 2(a) showing the typical core/shell structure of the black TiO2 [33]. The combination of X-ray diffraction (XRD) and Raman measurements has also confirmed the surface structural differences between black and white TiO2 nanomaterials. However, it should be noted that not all black TiO2 nanomaterials have the crystalline/disordered core/shell structure.

2.2. Surface Functional Group

Surface –OH groups exist in many TiO2 nanomaterials depending on the preparation method. As hydrogen is used in hydrogen thermal treatment and hydrogen plasma treatment, it is very possible to generate –OH groups on the surface of hydrogenated black TiO2 nanomaterials owing to the reduction effect of hydrogen. An increase in −OH groups has been detected in hydrogenated black TiO2 nanomaterials by X-ray photoelectron spectroscope (XPS) [22, 45]. Fourier transform infrared (FTIR) spectroscopy is another important technique to reveal the change in surface −OH groups by comparing the magnitude of the intensity of the peak corresponding to the −OH vibrational band [33, 34, 46]. In addition, 1H nuclear magnetic resonance (NMR) technique has also been used to analyze the surface –OH groups [33, 34].

The existence of –H groups on the surface of black TiO2 nanomaterials is very debatable. They have only been detected in a few cases of hydrogenated TiO2 nanomaterials [33, 47]. Wang et al. attributed the peak at 457.1 eV in the Ti 2p XPS spectrum of the hydrogenated black TiO2 nanocrystals to surface Ti–H bonds [33]. Zheng et al. found that hydrogenated TiO2 nanowire microspheres exhibited one shoulder peak at the lower binding energy side of the broader Ti 2p peak in the XPS spectrum and attributed it to the surface Ti–H bonds formed under hydrogen atmosphere [47]. Formation of surface Ti–H bonds was at the expense of surface Ti–OH groups [47]. Such groups do not exist on the surface of white TiO2 nanomaterials undoubtedly.

2.3. Defects

Oxygen vacancy is one of the most common defects existing in metal oxides including TiO2 nanomaterials [4850]. It has considerable influence on the activity and kinetics of the reactions proceeding on the surface of metal oxides [15, 4851]. Oxygen vacancies are undetectable in most white TiO2 nanomaterials, while oxygen vacancy is considered to be one of the feature defects in most black TiO2 nanomaterials. Electron paramagnetic resonance (EPR) or electron spin resonance (ESR) spectroscope [3740, 52], Raman spectroscope [9, 16, 32, 5355], and X-ray diffractometer [23, 32, 55] have been used to detect the oxygen vacancies present in black TiO2 nanomaterials.

Ti3+ defects do not exist in white TiO2 as all Ti ions usually present in the form of Ti4+. The presence of Ti3+ defects in black TiO2 nanomaterials is debatable. Based on XPS technique, some researchers reported the absence of Ti3+ defects in black TiO2 nanomaterials [22, 23, 34, 45, 46], while others argued the presence of Ti3+ defects derived from the reduction of Ti4+ ions [9, 16, 30, 53, 5558]. More commonly, Ti3+ defects in black TiO2 could be detected by EPR or ESR spectroscope [30, 33, 37, 41, 52, 55, 57, 59, 60].

Oxygen vacancies and/or Ti3+ impurities are believed to be highly related to the color change of the defect-rich TiO2 [14, 6166]. More importantly, oxygen vacancies and/or Ti3+ defects extend the photoresponse of TiO2 from UV to visible light region, which leads to high visible-light photocatalytic activity [14, 59, 6772]. Note that high density of crystal defects, especially bulk defects, may accelerate electron-hole recombination as defects can act as charge annihilation centers [57, 73, 74]. It is likely that surface oxygen vacancies are responsible for the enhanced photocatalytic activity [39, 70, 75], while bulk oxygen vacancies act as trap states and charge carrier recombination centers [76]. Studies also showed that surface Ti3+ defects could enhance hole trapping and thus facilitate the separation of photo-excited electrons and holes [57], while bulk Ti3+ sites acted as charge annihilation centers, leading to enhanced nonradiative recombination and shorter lifetime of electrons and holes [57, 60]. In addition, Ti3+ ions with oxygen vacancies can improve the electrical conductivity of TiO2 [73, 77], which may enhance the charge transport and charge-transfer reaction [73].

2.4. Electronic Band Structure

The introduction of defects and/or surface disorder could result in change in electronic and optical properties of black TiO2. Bandgap narrowing is a vivid demonstration of the change in band structure of black TiO2. The origin of the bandgap narrowing has been argued for years owing to the complexity of the surface defects/disorders and functional groups. One assumption is that Ti3+ species in the bulk TiO2 is responsible for the band gap narrowing [14, 59, 67, 68], and the oxygen-vacancy midgap states further enhance the light absorption for photon energy below the direct bandgap by indirect electron transitions [14]. Chen et al. thought that bandgap narrowing mainly originated from the disorder-induced midgap states rather than the Ti3+ species, greatly upshifting the valence band (VB) edge of black TiO2, and that the possible conduction band (CB) tail states arising from the surface disorder could only slightly narrow the bandgap (Figure 3(a)) [22, 34]. Naldoni’s group agreed with these analyses and further pointed out that oxygen vacancies could introduce localized states at 0.7–1.0 eV below the CB minimum of black TiO2 [23]. As shown in the schematic illustration of the density of states (DOS) of black and white TiO2 (Figure 3(b)), Wang et al. reported that the CB and VB tails slightly narrowed the bandgap by 0.8 eV and that the Ti–H bonds introduced the midgap electronic states at 0.92–1.37 eV below the CB minimum of black TiO2 [33]. In summary, the change in band structure is mainly attributed to the tailing of VB and/or CB, and the midgap states are induced by oxygen vacancies or −H groups [22, 23, 33, 39].

3. Fundamental Physicochemical Process in Photocatalysis

A simplified model representing the fundamental physicochemical process in photocatalysis is demonstrated in Figure 4. A typical process of photocatalysis involves three steps: light absorption, electron-hole separation, and surface reaction. Light with energy greater than the bandgap of TiO2 nanocrystal excites an electron from the VB to the CB; meanwhile, a positive hole will be left in the VB. In the case of anatase TiO2 with a bandgap of 3.2 eV, UV light with λ ≤ 387 nm is required for electron excitation. The electrons and holes that have been separated and transferred onto the surface of TiO2 can trigger redox reactions (pathways 1 and 2). For example, the electrons scavenged by O2 can yield superoxide radical anions, while the holes that react with H2O can produce hydroxyl radicals. These radicals can oxidize organic species, such as methylene blue [6], rhodamine B and salicylic acid [8], and methylene orange [9]. Thus, photocatalysis can be applied in degradation of pollutants [6, 8, 9], reduction of CO2 [10], and water splitting [1, 6]. However, electrons and holes may recombine on the surface (surface recombination, namely, pathway 3) or even in the bulk (volume recombination, namely, pathway 4) [74], which will compete with the desired redox reactions and thus greatly decrease the efficiency of the photocatalytic process. To increase the utilization of solar energy, the bandgap of TiO2 needs to be engineered to extend its optical absorption from UV to visible even to infrared light. On the other hand, the charge carrier separation must be exponentially enhanced in order to improve the photocatalytic efficiency and achieve the ultimate practical applications of TiO2 in photocatalysis.

4. Black TiO2 Nanomaterials as Visible Light-Active Photocatalysts

Black TiO2 nanomaterials have been synthesized by various methods including hydrogen thermal treatment, hydrogen plasma treatment, chemical reduction, chemical oxidation, and electrochemical reduction. Although those black TiO2 nanomaterials have similar appearance, their microstructures may differ owing to the differences in preparation methods and reaction parameters, and thus, black TiO2 nanomaterials from different research groups worldwide possess different physicochemical properties. Herein, we will give a few typical examples for each preparation method.

4.1. Black TiO2 Nanomaterials by Hydrogenation

Since the discovery of black TiO2 by Chen et al. in 2011 [22], hydrogenation or hydrogen reduction has become a powerful tool to synthesize black TiO2 nanomaterials. A variety of parameters, such as source of raw TiO2, hydrogenation time and temperature, H2 pressure, exposed crystal facet of TiO2, and even reactor materials, will affect the colorization, surface structure and groups, and photocatalytic performance of hydrogenated TiO2 nanomaterials [22, 23, 36, 40, 4547, 52, 73, 7680]. Note that the crystalline/disordered core/shell structure is only observed in a few hydrogenated black TiO2 nanomaterials [22, 23, 36, 78]. Chen et al. synthesized black TiO2 nanocrystals by high-pressure hydrogenation in a 20.0-bar H2 atmosphere at about 200°C for 5 days [22]. The raw TiO2 nanocrystals were prepared with a precursor solution consisting of titanium tetraisopropoxide, ethanol, hydrochloric acid, deionized water, and Pluronic F127 as an organic template, followed by hydrolysis ay 40°C for 24 h, solvent evaporation at 110°C for 24 h, and calcination at 500°C for 6 h in air [22]. The disorder-engineered black TiO2 nanomaterials contain two phases with a core/shell structure: a crystalline core and a disordered or amorphous shell (Figure 5). The crystalline phase of the black TiO2 maintained the anatase structure of the white raw TiO2 as evidenced by XRD analysis, whereas peak broadening and extra peaks besides the typical signals of anatase TiO2 were observed in the Raman spectrum of the black TiO2 owing to the disordered nature of the surface layer [22]. When used as photocatalysts, the degradation rate of methylene blue on the black TiO2 nanocrystals was found to be nearly 7 times that on the raw TiO2 nanocrystals, and the photocatalytic H2 production rate of the black TiO2 using 1 : 1 water-methanol solution under sunlight reached as high as 10 mmol·h−1 g−1 [22]. The high photocatalytic activity was ascribed to the substantially narrowed bandgap (experimentally ~1.54 eV) induced by the surface disorder, thus extending the optical absorption from UV to infrared region [22]. First-principle density functional theory calculation showed that two midgap states (centered at ~1.8 and ~3.0 eV corresponding to the VB and CB band tails, resp.) were created in the black TiO2 nanocrystals and thus well explained the origin of the change in the electronic and optical properties of black TiO2 nanocrystals [22].

Black rutile TiO2 nanowires have been prepared by hydrogenation in a tube furnace filled with ultrahigh purity hydrogen gas at temperatures ≥ 450°C for 30 min for photoelectrochemical water splitting [45]. Pristine rutile TiO2 nanowires were grown on a fluorine-doped tin oxide (FTO) glass substrate by hydrothermal method with titanium n-butoxide in aqueous hydrochloric acid solution at 150°C for 5 h, followed by annealing in air at 550°C for 3 h to increase the crystallinity of TiO2 nanowires and improve their contact to the substrate [45]. In this study, the authors found that the enhanced photoactivity of the black TiO2 nanowires was mainly attributed to the facilitated charge transport in black TiO2 and charge separation at the TiO2/electrolyte interface owing to the increased donor density or oxygen vacancies, while the visible light absorption only made a negligible contribution [45]. Further study showed that the electronic band structure of the black TiO2 nanowires was similar to that of the pristine TiO2 nanowires, and the color change was owed to the impurity/defect states in the bandgap of the black TiO2 nanowires caused by oxygen vacancies [45]. Similar enhancement in photoactivity of the hydrogenated anatase TiO2 nanotubes was observed in this study [45].

Liu et al. fabricated black anatase TiO2 nanotubes by anodic oxidation and hydrogenation [52]. In a typical synthesis, clean titanium foils were anodized in a two-electrode configuration with platinum gauze as the counter electrode in the electrolyte containing ethylene glycol (less than 0.2 wt.% H2O), deionized water (1 M), and NH4F (0.1 M) at 60 V for 15 min; black TiO2 nanotubes were then achieved by atmospheric pressure hydrogenation in H2/Ar (5%) flow at 500°C for 1 h [52]. The authors compared the effect of a series of factors on the photoactivity of the hydrogenated TiO2. It was found that atmospheric pressure (H2/Ar) hydrogenated anatase TiO2 nanotubes had more oxygen vacancies, while high-pressure (20.0 bar, 500°C for 1 h) hydrogenated anatase TiO2 nanotubes possessed more Ti3+ species [52]. Photocatalytic experiments demonstrated that high-pressure hydrogenated anatase TiO2 nanotubes exhibited a high H2 evolution rate of 7 μmol·h−1 cm−2 without any cocatalysts, while atmospheric pressure hydrogenated anatase TiO2 nanotubes had negligible H2 evolution [52]. For reference, rutile nanorods hydrogenated in either atmospheric pressure or high-pressure hydrogen showed extremely small H2 evolution rate [52]. It is believed that different polymorphs using anatase and rutile as examples have different defect formation behaviors upon reductive treatments [66, 81, 82] and thus lead to different photoactivities of treated TiO2 nanomaterials.

Recently, ordered mesoporous black anatase TiO2 was prepared through an evaporation-induced self-assembly method combined with an ethylenediamine encircling process, followed by atmospheric pressure hydrogenation at 500°C for 3 h under H2 flow [78]. Figure 6(a) displays the schematic synthesis process. Interestingly, the ordered mesoporous TiO2 prepared with ethylenediamine turned into a black color after hydrogenation (Figure 6(b)), whereas only gray TiO2 was obtained with the porous TiO2 synthesized without ethylenediamine [78]. This suggested that the original surface functionalities and surface structural defects may be the very important factors determining the colorization or types of defects or electronic band structure of the hydrogenated TiO2 nanomaterials. The VB XPS analysis found that the VB maximum energy for the black TiO2 blue-shifted towards the vacuum level at ~1.6 eV owing to the possible Ti3+ species, narrowing the bandgap to ~2.80 eV consistent with the experimental value of 2.82 eV [78]. The resultant ordered mesoporous black TiO2 showed an extended photoresponse from UV light to visible and infrared light regions and thus exhibited a high photocatalytic hydrogen evolution rate of 136.2 μmol·h−1, which approached two times that (76.6 μmol·h−1) of the pristine ordered mesoporous TiO2 [78].

4.2. Black TiO2 Nanomaterials by Hydrogen Plasma

Hydrogen plasma technology has attracted increasing interest owing to its effectiveness in engineering surface-disordered TiO2 nanomaterials with a typical crystalline/amorphous core/shell structure [33, 39, 57, 60]. Ti3+ species and oxygen vacancies were reported to be the primary defects in some cases [57, 58], while Ti–H groups and oxygen vacancies were believed to be the dominant defects in other cases [33, 39].

H-doped black titania (TiO2−xHx) with high solar absorption (~83%) was converted from commercial TiO2 (Degussa P25) in a thermal plasma furnace by hydrogen plasma for 4–8 h at 500°C with the plasma input power of 200 W [33]. The TiO2−xHx presented a crystalline/amorphous core/shell structure, and the black coloration of TiO2−xHx was possibly caused by the defects including H doping, oxygen vacancies and surface hydroxyl groups [33]. As discussed above in Section 2.4, the tailing effect of VB and CB and the midgap states caused by H doping contributed to the Vis-NIR absorption of the black TiO2−xHx and thus highly enhanced photocatalytic activity [33]. The TiO2−xHx showed much higher solar-to-electron efficiency in both photocatalytic water splitting (Figure 7(a)) and degradation of methyl orange over pristine TiO2 (Figure 7(b)) and demonstrated a high cycling stability (Figure 7(c)) [33].

Black TiO2 nanocrystals with the typical crystalline/amorphous core/shell structure were prepared from hydrogen plasma treatment of commercial TiO2 nanoparticles (Degussa P25) by Yan et al. [57]. Surprisingly, the authors found that slightly hydrogenated TiO2 nanocrystals without visible coloration exhibited enhanced photoactivity in both photocatalytic degradations of methylene blue and reduction of CO2 with H2O, while gray and black TiO2 showed worse photoactivity over pristine TiO2 [57]. It was proposed that improved photocatalytic performance of slightly hydrogenated TiO2 could be attributed to the higher ratio of trapped holes (O centers) and a lower recombination rate induced by the increase of surface defects, while the highly concentrated bulk defects in gray and black (overhydrogenated) TiO2 acted as charge recombination centers, leading to worse photoactivity [57].

4.3. Black TiO2 Nanomaterials by Chemical Reduction
4.3.1. Aluminum Reduction

Huang’s group has reported on the synthesis of a series of black TiO2 nanomaterials by aluminum reduction [9, 32, 37, 38, 41, 53]. The typical crystalline/amorphous core/shell structure was observed in almost all Al-reduced black TiO2 [9, 32, 37, 38, 41], except for the black TiO2 nanotubes [53]. Also, Ti3+ and oxygen vacancy defects were commonly detected in Al-reduced black TiO2 [9, 32, 37, 38, 41, 53]. In a typical synthesis of the Al-recued black TiO2, aluminum and pristine TiO2 (Degussa P25) were placed separately in a two-zone tube furnace (Figure 8(a)); the pressure in the tube was controlled at a base pressure below 0.5 Pa, and then in order to trigger the reduction reaction, aluminum was heated at 800°C while TiO2 was heated at 300–500°C for 6 h [32]. As shown in Figure 8(b), black TiO2 nanoparticles can be produced on a large scale with aluminum reduction method. A unique crystalline/amorphous core–shell structure was observed on all Al-reduced TiO2 prepared at different temperatures, and the thickness of the disordered outer layer increased with the Al-reduction temperature (Figures 8(c), 8(d), 8(e), and 8(f)) [32]. The black TiO2 absorbed ~65% of the total solar energy by improving visible and infrared absorption and thus exhibited markedly high photoactivity in both photocatalytic water splitting and degradation, superior to the pristine TiO2 (~5% solar energy absorption) [32].

4.3.2. CaH2 Reduction

Reduction of TiO2 (rutile in [83, 84]) by CaH2 usually generates black Ti2O3 rather than black TiO2 [83, 84]. Recently, black TiO2 was prepared from Degussa P25 by CaH2 reduction at 400°C [42]. The black TiO2 had a crystalline/amorphous core/shell structure with abundant oxygen vacancies, which led to a high solar absorption (~81% solar energy absorption) and significantly enhanced photocatalytic organic degradation and water-splitting performance [42].

4.3.3. Magnesium Reduction

Recently, Sinhamahapatra et al. developed a new method to synthesize black TiO2 with magnesium as the reductant [11]. Typically, well-mixed reactant of commercial TiO2 and magnesium powder was placed in a tube furnace and then heated at 650°C for 5 h in the flow of 5% H2/Ar; the product was stirred for 24 h in 1.0 M HCl and then washed with sufficient amount of water to remove the acid and dried at 80°C [11]. A small amount of anatase was transformed into rutile during the reduction process [11]. The maximum hydrogen production rates were 43 mmol·h−1 g−1 and 440 μmol·h−1 g−1, along with remarkable stability under full solar wavelength light and visible light irradiation, respectively [11]. This outstanding activity can be correlated with the extended absorption in visible light, perfect band position, the presence of an appropriate amount of Ti3+ species and oxygen vacancies, and slower charge recombination [11].

4.3.4. NaBH4 Reduction

3D mesoporous black TiO2/MoS2/TiO2 (MBT/MoS2/MBT) nanosheets were prepared by ball milling and subsequent NaBH4 reduction at 350°C for 1 h under an Ar atmosphere as shown schematically in Figure 9 [85]. The introduction of the TiO2-MoS2 heterojunction and the Ti3+ species narrowed the band gap of TiO2, leading to the excellent activity in photocatalytic degradation of methyl orange and water splitting for H2 evolution under visible-light irradiation [85]. The H2 production rates were 0, 0.13, 0.32, and 0.56 mmol·h−1 g−1 for mesoporous TiO2 (MT), mesoporous black TiO2 (MBT), (MT/MoS2/MT), and (MBT/MoS2/MBT), respectively [85].

4.3.5. Lithium Reduction

Zhang et al. reported that black rutile TiO2 can be achieved by soaking rutile TiO2 nanomaterials in a Li-containing ethanediamine solution [86, 87]. Typically, 14 mg of metallic Li foils were dissolved in 20 mL of ethanediamine under dry conditions; then, 200 mg of TiO2 nanocrystals (Degussa P25) was immersed into the ethanediamine solution for 6 h with continuous stirring; 1 M HCl was used to consume the excess Li or electrons when the reaction was complete; finally, the product was rinsed with deionized water several times and dried in vacuum oven at room temperature [86]. Note that the Li-assisted reduction is phase selective: rutile phase is reduced into black TiO2 while anatase phase is well-maintained [86]. This offers us an opportunity to create abundant order/disorder junctions at the surface by controlling the phase composition in pristine TiO2 for highly efficient photocatalytic hydrogen generation [86]. The order/disorder/water junction was believed to efficiently internally drive the electron/hole separation through type-II bandgap alignment and to trigger a strong hydrogen evolution surface reaction [86]. Furthermore, Zhang et al. found that the “crystal-deficient” layers on the surface of the rutile TiO2 nanowires increased the conductivity by 50 times, which increased the electron diffusion length to ~20 μm and overcame the charge collection limitation at the solid/liquid interface for efficient conversion of solar energy to chemical energy [87]. These studies highlight the importance of controlling the surface localization of defects and the solid/liquid interface towards enhanced photoactivity over TiO2 photocatalysts [8688].

4.4. Black TiO2 Nanomaterials by Chemical Oxidation

Xin et al. prepared black anatase TiO2 with a crystalline/amorphous core/shell structure by chemical oxidation method [30]. Typically, a yellowish gel was first obtained by reacting TiH2 and H2O2 for 12 h; then, the gel was diluted using ethanol, the pH of the mixture was adjusted to 9.0 by NaOH, and NaBH4 as an antioxidant was added to the resulting mixture; after the solvothermal treatment at 180°C for 24 h, the collected sample was washed with HCl, water, and ethanol; light blue TiO2 nanocrystals (TiO2−x) were obtained after the precipitate was dried in vacuum for 12 h; finally, postannealing treatment was carried out at 300–700°C for 3 h under a nitrogen flow [30]. The annealing temperature was found to be a crucial factor affecting the color of the as-prepared TiO2 nanocrystals [30]. Light brown, brown, black, dark brown, and shallow dark brown TiO2 nanocrystals were obtained at 300, 400, 500, 600, and 700°C, respectively [30]. It is found that the Ti3+ species initially increased with annealing temperature up to 500°C and then decreased with further temperature rise [30]. The black TiO2−x prepared at 500°C had the highest Ti3+ concentration and thus exhibited the highest photocatalytic activity [30].

Similarly, Grabstanowicz et al. reported on black rutile TiO2 that was prepared by oxidizing TiH2 in H2O2, followed by calcinations in Ar gas [59]. The Ti3+ concentration reached as high as one Ti3+ per ∼4300 Ti4+ in the black rutile TiO2, and thus, it exhibited remarkably enhanced visible-light photocatalytic degradation on organic pollutants in water [59]. Xin et al. fabricated black brookite TiO2 single-crystalline nanosheets by hydrothermal reaction with TiH2 and H2O2 as the Ti source and oxidant, respectively, followed by postannealing treatment at 500°C (T500 in Figure 10(a)) [89]. The black TiO2 showed drastically enhanced visible-light absorption with a significantly narrowed bandgap of 2.10 eV (Figure 10(a)) owing to the introduction of bulk Ti3+ defects [89]. When used as photocatalysts, the black TiO2 exhibited the highest CO2 reduction rate (11.9 μmol·g−1 h−1 for CH4 and 23.5 μmol·g−1 h−1 for CO) (Figure 10(b)) [89].

4.5. Black TiO2 Nanomaterials by Electrochemical Reduction

The electrochemically reduced black TiO2 often possessed abundant Ti3+ species and oxygen vacancies [17, 56, 90, 91]. The reported black TiO2 nanomaterials with a nanotube morphology and an anatase phase are prepared by electrochemical reduction in ethylene glycol electrolytes [56, 90, 91]. However, it should be noted that electrochemically reduced black TiO2 in ethylene glycol electrolytes was not stable [56, 91], because glycerol has a higher viscosity making it difficult for the protons to insert into TiO2 [91]. It is worth noting that the electrochemically reduced black TiO2 nanotubes were recently found unstable in air [17].

Xu et al. reported on the electrochemically hydrogenated black TiO2 nanotubes [90]. The pristine anodic TiO2 nanotubes were prepared at 150 V for 1 h in an ethylene glycol electrolyte containing 0.3 wt.% NH4F and 10 vol.% H2O with carbon rod and Ti foil as the cathode and anode, respectively [90]. After the pristine TiO2 nanotubes were annealed in air at 150°C for 3 h and then 450°C for 5 h, the electrochemical reduction was performed at 5 V for 5 to 40 s in 0.5 M Na2SO4 aqueous solution at room temperature with the TiO2 nanotubes as the cathode and Pt as the anode to achieve black TiO2 nanotubes [90]. The surface oxygen vacancies were considered to contribute to the substantially enhanced electrical conductivity and photoactivity [90].

Similarly, Li et al. synthesized black anatase TiO2 nanotubes by electrochemical reduction [91]. In their study, a so-called “activation” step was adopted in order to obtain stable black TiO2 nanotubes, where anodization was carried out in an ethylene glycol solution of 0.2 M HF and 0.12 M H2O2 at 60 V for 30 s before the electrochemical reduction synthesis of black TiO2 nanotubes at a cathodic voltage of −40 V for 680 s in an ethylene glycol solution of 0.27 wt.% NH4F [91]. The authors proposed the doping mechanism as follows:

H+ ions were likely driven inside the TiO2 under the cathodic field [91]. As shown in Figure 11, black TiO2 nanotubes demonstrated the highest photocatalytic activity in rhodamine B degradation experiment, though the as-anodized TiO2 nanotubes were rich in oxygen vacancies [91].

4.6. Black TiO2 Nanomaterials by Other Methods
4.6.1. Water-Plasma-Assisted Synthesis

Panomsuwan et al. reported on the water-plasma-assisted synthesis of black titania spheres (H-TiO2−x) with efficient visible-light photocatalytic activity [44]. The H-TiO2−x was composed of a mixture of rutile, anatase, and oxygen-deficient phases (e.g., Ti10O19, Ti5O9, and Ti3O5). The abundant oxygen vacancies and Ti3+ species with the presence of Ti2+ species resulted in a narrowed bandgap of 2.18 eV [44]. Under visible-light irradiation for 180 min, the methylene blue is almost completely degraded in the presence of H-TiO2−x (90%), whereas it is degraded only by 18% in the presence of P25 [44].

4.6.2. Nitrogen Doping

Wei’s group focused on the synthesis of crystalline/disordered core-shell black anatase TiO2 (TiO2@TiO2−x) by a one-step calcination in N2 [43, 92]. Typically, two mixture solutions were prepared: one containing tetrabutyl titanate, urea, and ethanol absolute and another containing hydrochloric acid, deionized water, and ethanol absolute; the latter was added dropwise to the former solution and stirred until white colloid was formed; the mixture was placed in a water bath at 35°C for 30 min and then stirred magnetically for 2 h; the TiO2@TiO2−x was obtained by annealing at 550°C for 3 h in a nitrogen atmosphere [43]. Oxygen vacancies and nitrogen species were detected in TiO2@TiO2−x which explained its narrowed bandgap and high visible light photocatalytic degradation performance on methyl orange [43]. The authors also investigated the effect of urea concentration on the structure and photocatalytic activity of the black TiO2 and found that a lower urea concentration triggered the largest amount of oxygen vacancies [43].

4.6.3. Electrochemical Oxidation

Defective black anatase TiO2 nanotubes were synthesized via two-step anodization on Ti foil in ethylene glycol containing 0.25 wt.% NH4F and 2 vol.% distilled water at 60 V for 10 h, followed by calcination in the air (Figure 12) [54]. The black TiO2 with controllable level of defects exhibited a high photocatalytic activity under visible light [54]. Mechanistic analysis and characterization results indicated that oxygen vacancies were formed in an oxygen-deficient environment during the anodization process and accounted for the high photon-absorbance of the black TiO2 throughout the visible-light region [54].

4.6.4. Ionothermal Synthesis

Black Ti3+-doped anatase TiO2 was synthesized by treating metal Ti in an N-N-dimethylformamide solution containing 1-methyl-imidazolium tetrafluoroborate (ionic liquid), lithium acetate, and acetic acid in an autoclave at 200°C for 24 h [55]. The ionic liquid enriched with fluorine and acetic acid play key roles in dissolving and oxidizing the Ti foil, respectively [55]. The in situ generated H2 from the reaction between Ti foil and acetic acid could be strongly adsorbed onto the surface of TiO2 and dissociated into H atom to form H-TiO2-x disordered layer [55]. The Ti3+-rich black TiO2 exhibited high activity in photocatalytic degradation of organic pollutants under visible light (λ > 420 nm) and also a high hydrogen evolution rate of 0.26 mmol·h−1 m−2 in water splitting under simulated solar light [55].

4.6.5. Laser Irradiation

Black TiO2 nanospheres were fabricated by laser irradiation on suspended solution containing 20 mg anatase TiO2 nanospheres and 1 mL distilled water [93]. Phase transformation from anatase to rutile was observed when laser irradiation was performed more than 15 min, and black TiO2 was obtained after 120 min laser irradiation [93]. The degradation ratio of rhodamine B with black TiO2 nanospheres could reach up to 33% under green light for 5 h, while the pristine and P25 showed no degradation ability on rhodamine B [93]. The high photoactivity was attributed to the Ti3+ defects and disordered surface layer which resulted in a narrowed bandgap of 2.2 eV [93]. Recently, black amorphous TiO2 film was achieved by pulsed laser deposition at 100°C for 10 min under vacuum condition using a commercial TiO2 target and a KrF excimer laser at a repetition rate of 2 Hz with a laser fluence of 2 J·cm−2 [94]. This black amorphous TiO2 film was deposited on a predeposited crystalline TiO2 film to construct a bilayer structure similar to the crystalline/amorphous core/shell structure of black TiO2 nanoparticles in order to create a simpler model to elucidate the working mechanism of black TiO2 nanomaterials in many applications [94]. Metallic conduction was achieved at the crystalline/amorphous homointerface via electronic interface reconstruction [94]. This points to a research direction that may partly eluciate the high performance of black TiO2 in many applications.

4.6.6. Proton Implantation

Liu’s group applied proton implantation method to the synthesis of black TiO2 nanotubes [95]. Proton implantation was carried out at an energy of 30 keV and a nominal dose of 1016 ions·cm−2 using a Varian 350 D ion implanter [95]. While the ion implantation on a (001) surface plane of an anatase crystal led to a low H2 production efficiency, implantation of TiO2 nanotubes markedly enhanced hydrogen evolution due to the length effect [95]. That is, a synergistic interaction between the implanted upper part of the TiO2 nanotube, acting as light absorber, and the intact lower part, acting as catalytically active center, was proposed [95].

5. Summary and Prospective

TiO2 photocatalyst as an ideal model for the investigation of photocatalysis in a variety of areas has attracted enormous attention over the past decades, especially on the water splitting for hydrogen production, pollutant removal for environmental protection and CO2 reduction for solar fuels. These aspects are significant to the sustainable development of the economy and human society. However, the solar energy conversion efficiency is quite low owing to the limited optical absorption of TiO2 only to UV spectrum and the rapid electron-hole recombination. Structural engineering that mainly leads to surface structure change of TiO2 with color change from whiteness to blackness is a promising strategy. This strategy has the potential to take care of both sides, that is, black TiO2 nanomaterials often exhibit a large absorption of visible light and highly enhanced charge separation. Many synthetic approaches have been developed to synthesize surface-structure-engineered black TiO2 nanomaterials since the report of the hydrogenated black TiO2 crystals in 2011. Despite these progress, two main challenges are present in terms of synthesis and applications of black TiO2. (i) Optical absorption of visible light, at present, does not mean their successful conversion to solar fuels; increasing the effective utilization of visible light is still a challenge for black TiO2. (ii) Scalable synthesis of black TiO2 with highly controllable quality is of vital importance for its practical applications. On the other hand, there are still quite a few open questions: why black TiO2 nanomaterials have much higher photocatalytic activity over normal white TiO2 nanomaterials (however, note that not all black TiO2 nanomaterials exhibited enhanced photoactivity); what on earth primarily triggers the optical absorption of black TiO2 nanomaterials to visible light; what are the individual roles of different types of defects in black TiO2 nanomaterials; how the defect-rich surface region in black TiO2 nanomaterials affects the charge transfer and photoactivity; etc. The answers to these fundamental questions vary with preparation method, properties of pristine TiO2, et al. This may be owing to the high sensitivity of the optical and electronic properties of TiO2 to its surface structure. The huge differences in the surface structures (including surface groups, types and quantities of defects, and heteroatom contamination) among different black TiO2 make the fundamental understanding even harder. Therefore, novel strategies along with new technologies that can precisely control and probe the surface structural evolution during the preparation process are highly desired. TiO2 with quantitatively controlled surface defects may be a good model to reveal the underlying working principle of black TiO2 nanomaterials in photocatalysis. The interface between the crystalline core and the disordered surface layer may also be an important consideration for understanding the basic physicochemical properties of black TiO2, but less attention has been paid to that. On the other hand, the interaction between the reactant compounds/ions and the surface of black TiO2 nanomaterials, especially the effects of their surface states on the adsorption and activation of the targeted reactants, also calls for more attention as different photocatalytic processes need slightly different reaction environments which are highly related to the surface structures. We hope this review can inspire more work to advance the understanding and development of the black TiO2 in photocatalysis.

Conflicts of Interest

The authors declare that they have no conflict of interest.


Xiaodong Yan thanks the funds provided by the University of Missouri-Kansas City, School of Graduate Studies.