1. Reactive Oxygen Species in TiO2 Photocatalysis
ROS can be considered primary intermediates of photocatalytic reactions with these four recognized as the main ones: hydroxyl radical (·OH), superoxide anion radical (·O
2−), hydrogen peroxide (H
2O
2), and singlet oxygen (
1O
2)
[1][2]. ROS seem to form mainly from the interaction of the VB hole with molecules (such as H
2O) or species, oxidizing the latter and typically resulting in ·OH centers
[3][4][5]. This could also happen in hole-trapping processes in TiO
2, such as at bridging O
2−, resulting in the formation mainly of ·O
− (“deprotonated ·OH”)
[6][7][8]. Because of the high potential barrier of free ·OH for desorption, adsorbed ·OH is considered more favorable and is usually equated to trapped holes due to the adsorption–desorption equilibrium
[1][6].
A detailed summary of generating the four major ROS on the TiO
2 surface can be viewed in terms of bridging and terminal OH sites
[3]. The reactions occurring at the anatase and rutile are differentiated by the arrow lines (double lines are restricted to anatase), whereas broken lines refer to adsorption/desorption. At the bridged OH site, a photogenerated hole attacks the O
2− bridge (step a), forming Ti-O· and Ti-OH (step b), which can be reversed by recombining with an electron from the CB. Some surface-trapped holes at the anatase can be released as ·OH into the solution. At the rutile surface with its suitable distance between adjacent Ti surface atoms, a different scenario occurs. Once another hole is formed in the same trapped hole-containing particle, the hole could migrate and interact with the existing hole resulting in a peroxo-bridged structure at the surface (step c). Further reactions of these structures could then generate other ROS
[1][3].
At the terminal OH site, a photogenerated electron can interact and is trapped at the Ti
4+ site, transforming it to Ti
3+ (step b)
[9]. The trapped electron in the Ti
3+ could then reduce oxygen to form an ·O
2− (adsorbed) (step d) (which, with further reduction, could become an adsorbed H
2O
2 or as (Ti)-OOH (step c)). The H
2O
2 that is adsorbed could also be reoxidized to produce an adsorbed ·O
2−, which can be desorbed to return to the initial state (step a). As the peroxo bond needs to be dissociated, the production of ·OH is highly unlikely when the adsorbed H
2O
2 is being reduced
[1][3]. These schemes also show a sensible explanation for the influence of the adsorption of H
2O
2 in forming ROS, which has been well-considered for increasing the photocatalytic performance of TiO
2.
Adsorbed ROS can also have a more direct impact on photocatalytic performance
[1]. Anatase and rutile show different reactivity towards forming ·OH and ·O
2− [10][11], likely due to their H
2O
2 adsorption
[8] in addition to their band edge alignment. One-step oxidation of H
2O
2 produces ·O
2−, which is more remarkable for anatase, whereas one-step reduction produces ·OH, which is only observed for rutile or rutile-containing forms and is believed to be due to the structure of the adsorbed H
2O
2 on rutile vs. anatase
[12].
2. Nanomorphologies and Structural States of TiO2
Due to its wide applicability, TiO
2 has been produced via different means, with the resulting TiO
2 structural polymorphs—i.e., anatase, rutile, or brookite, among others
[13]—and morphology being highly influenced by the preparation method. The different TiO
2 morphologies add to the variety of properties and performance exhibited by TiO
2. In addition to bulk TiO
2 [14][15][16][17], in recent decades, TiO
2 nanomaterials of various morphologies have also been developed, resulting in the current plethora of TiO
2 nanomorphologies (
Figure 1). These have been synthesized using different means for various targeted applications, achieving a range of photocatalytic efficiencies. Note that some morphologies are preferably prepared using certain procedures (e.g., sol-gel method for nanopowders and anodization for nanotubes), whereas some procedures (e.g. hydrothermal synthesis) can be used and modified to produce various morphologies (such as nanospindles, nanorhombus, nanorods, or nanosheets).
Figure 1. Plethora of TiO2 nanostructures. This includes free-standing nanomaterials, such as TiO2 dendritic nanospheres, nanospheres, nanoshells, nanofibers, and nanoflowers, and nanostructures grown on bulk substrates, such as nanotubes, nanolace, and nanotrees. Nanomaterials present increased catalytic activity due to their interesting properties (increased surface area, enhanced charge separation, light absorption/harvesting), which are desirable in catalytic TiO2 applications. For image reprint/adaptation, please cite the original source: Querebillo, C.J. A Review on Nano Ti-Based Oxides for Dark and Photocatalysis: From Photoinduced Processes to Bioimplant Applications. Nanomaterials 2023, 13(6), 982; https://doi.org/10.3390/nano13060982. Copyright 2023 Querebillo, C.J.
The morphologies of nanosized titanium dioxide can be grouped according to their dimension classifications: zero-dimensional (0D) includes nanopowders, nanocrystals, and quantum dots (QD); one-dimensional (1D) includes nanowires, nanofibers, nanotubes, and nanorods; two-dimensional (2D) includes nanosheets; and three-dimensional (3D) includes nanotube arrays. Mixed morphologies, such as nanosheets with QDs, also exist.
Many studies on photocatalysis have been conducted on nanoparticle suspensions
[18][19][20], which are inconvenient and result in practical difficulties
[21] because complete photocatalyst recovery is challenging. Therefore, studies have resorted to photocatalyst immobilization
[22][23][24][25], which, however, requires catalysts of high activity. TiO
2 of different nanomorphologies were used, evaluated, and/or compared in terms of performance
[26][27][28]. The photocatalytic performance reported, usually measured by dye degradation rate, varies and depends on the experimental setup/condition, such as the illumination and probe used, though the typical pseudo-first-order rate constant
k is within the 10
−3–10
−1 min
−1 range. Therefore, studies also sometimes include a reference TiO
2, such as commercially available P25 for benchmarking. Nevertheless, from this summary, one also sees that generally, some improvement in photocatalytic performance is brought about by nanomorphology based on the obtained
k values being mainly in the 10
−1–10
−2 min
−1 range for systems with varied morphology, which are at least one order of magnitude better than the usual for those with nanopowders (
k~10
−2–10
−3 min
−1).
Post-synthesis heat treatment (calcination/annealing) mainly dictates the structural state. Heat treatment can be performed to transform the amorphous state to rutile (≥600 °C) or anatase (300 °C to 500 °C) or to transform anatase to the more stable rutile
[29]. The crystalline state also influences photocatalytic activity. It is commonly agreed that anatase is better than the other states (such as rutile) due to its higher surface affinity and slower recombination rate. However, mixed states (such as the case of P25) also exhibit good photocatalytic activity, though the surface crystalline state seems to play a more crucial role in such cases of mixed states due to the fact that photocatalytic reactions take place at the surface
[30].
3. Photocatalytic Disinfection Using TiO2 Nanostructures
The photocatalytic ROS production on TiO
2 NPs, while it may pose some health risks, can also provide benefits. As early as 1985, the photocatalytic microbicidal effect of TiO
2 has been reported by Matsunaga et al.
[31]. More studies have then been carried out on the bacteria-killing action of TiO
2 [32][33][34][35]. Maness et al.
[32] attribute this effect to the lipid peroxidation in the microbe (in their case, in
E. coli) due to the photocatalytic oxidative property of TiO
2 NPs. Upon initiation of lipid peroxidation by ROS, propagation can happen via the generation of peroxy radical intermediates, which can also react with other lipid molecules. Superoxide radical could also be involved, as it can also be photogenerated on TiO
2. This can react with an intermediate hydroperoxide to form new reaction chains that can go through the damaged cell membrane. Once the cell wall is broken down, TiO
2 NPs themselves could also possibly directly attack the cell membrane
[32]. It is important to remember though that the microbe’s response to photocatalytic disinfection action can also be influenced by its level of protective enzymes against oxidative stress
[34].
It is generally accepted that the photocatalytic antibacterial properties of TiO
2 are due to its ROS formation, whereby ·OH is thought to play a crucial role
[34]. Yet, novel TiO
2-based materials also point to the role and use of other ROS. The development of different TiO
2 nanostructures paves the way for advancements in photocatalytic disinfection on TiO
2, and some examples, also in relation to the formation of ROS, are presented here.
Nanocomposites made from TiO
2 NPs on Si nanostructured surfaces can be used as antibacterial surfaces for dental and orthopedic implants, and the TiO
2 NPs themselves can be spray-coated to surfaces for disinfection of microbes upon irradiation
[11]. The nanostructures are said to rupture the bacterial cell wall, whereas the ROS from TiO
2 NPs can oxidize organic matter (such as bacteria) to prevent bacterial growth
[36], which could eventually form biofilms. On the other hand, the free radicals photogenerated on TiO
2 can also disrupt and destroy biofilms. This is important since killing bacteria within a biofilm is quite challenging; the biofilm shields the bacteria from antibiotics, antibodies, and immune cells
[37][38]. Using TiO
2 exposed to UVA, the biofilm formation of
P. aerigunosa was inhibited via ROS attack to disrupt the bacterial cell membrane (disabling bacteria to form biofilm), and the ε-poly(L-lysine) of the cells already in the biofilm was weakened
[11]. Through a plasma electrolytic oxidation process, Nagay and coworkers produced N- and Bi-doped TiO
2 coatings that kill bacteria because of the generation of ROS upon visible light exposure
[39][40].
Efforts to extend photosensitization with TiO
2 were performed by forming composites with materials such as MoS
2 and
l-arginine and/or doping with Yb and Er
[39][41]. However, the broader spectral range or near-IR sensitization is usually brought about by the additional component (and not by TiO
2). On the other hand, other works utilize the photogenerated charge carriers from TiO
2 and enhance the catalytic effect by the addition of other components for antibacterial purposes. For example, TiO
2 combined with graphdiyne (GDY) was synthesized into nanofibers by electrostatic force to produce ROS and prolong the antibacterial effect. When exposed to light, electrons and holes are generated on the TiO
2 and GDY surface, with the photogenerated electrons of TiO
2 being easily transferrable to the GDY surface. There, ·OH and ·O
2− are formed because they can react easily with water and O
2. The extended lifetimes of the charge carriers enhance ROS generation and the resulting bactericidal effect. Overall, these processes inhibit the methicillin-resistant
S. aureus (MRSA) biofilm formation and promote the regeneration of bone tissues
[11][39][42].
In general, the photocatalytic disinfection by TiO
2 nanocomposite antimicrobial coatings entails the incorporation of inorganic metals/nonmetals (such as Ag, Cu, Mn, P, Ca, and F) and/or 2D materials (graphydiyne, MXenes, and metal–organic frameworks, etc.) into TiO
2 to control the porosity of the surface, crystallinity, charge transfer, and disinfecting property against critical pathogens, such as
S. aureus and
E. coli but also H1N1, vesicular stomatitis virus (the safe surrogate virus for SARS-CoV-2), and the human coronavirus HCoV-NL63
[11]. Such light-catalyzed coatings could prevent microbes from reactivating to completely destroy them, and with the high mobility of ROS in the air, airborne microbes could also be targeted
[43]. The high interest in the inactivation and disinfection of coronaviruses emerged recently due to the recent pandemic. Some of these studies were on the photocatalytic disinfection of coronavirus using TiO
2 NP coatings, with the mechanism attributed to their generation of ROS
[11][44][45]. For interest in antimicrobial coatings, readers are referred to Kumaravel et al.
[11].
4. Efforts to Improve the Photocatalytic Efficiency of TiO2 Nanomaterials
The photocatalytic activity of TiO2 nanomaterials improves by resorting to different morphologies. Their nanosize alone increases the surface area, providing more active sites. Additionally, due to the interesting properties afforded by its size, the use of nanomaterials also improves photocatalytic efficiency by enhancing charge separation and light harvesting and increasing the surface-to-bulk ratio. These improvements are presented in this section.
4.1. Enhanced Charge Separation
The increased photocatalytic performance of TiO
2 nanoparticles cannot be attributed to the increased specific surface area alone but also to the
increase in the surface-to-bulk ratio with decreasing particle size. The latter results in shorter diffusion pathways that the charge carriers have to traverse to reach the surface, which is the photocatalytic reaction site
[46]. Adding adsorbed species further provides electron/hole scavengers that could also improve the charge separation
[46][47][48]. The decrease in the size though also blueshifts the TiO
2 absorption edge
[49] and could result in unstable NPs
[50]. Therefore, an optimized size is needed to balance the properties in terms of charge separation, light absorption, and stability. During thermal treatment, which is typically needed for good crystallinity and increased photocatalytic performance, aggregation could occur, and this can be prevented by preparing highly-dispersed TiO
2 clusters (such as those synthesized with zeolites
[51]), which also shows high photocatalytic activity. These spatially separated TiO
2 species were also prepared as single-site catalysts that also show high photocatalytic electron–hole pair reactivity and selectivity
[6]. The high photocatalytic reactivity observed for highly dispersed TiO
2 species is attributed to the highly selective formation of a longer-lived (up to µs), localized charge-transfer excited state compared to that of bulk TiO
2 (ns)
[46].
The aggregate formation is not always disadvantageous. The mechanisms portrayed for a single semiconductor NP are simplified, and a more accurate representation would consider that TiO
2 NPs have the tendency to self-aggregate in aqueous solution to form a 3D framework. This happens by aligning their atomic planes with each other, allowing for efficient charge carrier transport without interfacial trap interferences in a so-called “antenna effect”. Through this, the photogenerated excitons in a nanoparticle will be transported throughout the network until they get trapped individually in a suitable site (e.g., via a redox reaction with an adsorbed electron acceptor/donor on one particle in the network). Charge carriers that are not trapped continue to traverse through the network until they themselves react. Therefore, through forming 3D aggregates, better electron mobility is achieved for TiO
2 particles
[6][52].
When the aggregates align, they act as if they are an array of nanowires that facilitate efficient CT throughout the network. In fact, 1D morphologies, such as TiO
2 nanowires, nanotubes, and nanorods, are hailed for their efficient electron transport since the photoexcited charge carriers could move along the length, increasing delocalization and resulting in long diffusion lengths (>200 nm), which delays the charge recombination and prevents electrons from residing in traps
[53][54][55][56][57][58]. For example, using TiO
2 nanotube arrays (TNAs) for PEC water splitting can improve the photocatalytic efficiency and result in a photocurrent of up to 10 times since loss of photogenerated electrons is prevented with the electrons being able to diffuse along the tube towards the collecting substrate. The improved performance using TNA is said to not only come from enhanced charge separation and better electron transport due to the orderly arrangement
[53][59][60][61] of this 3D structure but also due to better light harvesting
[46][53].
Meanwhile, the photocatalytic activity of pure TiO
2 nanomaterials is considerably small and requires mostly UV light due to its bandgap. Hence, various attempts were conducted to improve its efficiency and increase its absorption range.
Doping is one widely used approach, and this has been performed with transition metal ions, such as Fe
3+, which increased the efficiency for pollutant photodecomposition likely due to the fact that it decreases the
Eg and broadens the absorption range to the visible region
[6][62]. Oxygen doping in TiO
2 interstices can also improve photocatalytic performance by enhancing charge separation efficiency. For the photodegradation of methyl orange, O-doped or oxidative TiO
2 showed a 2~3.7 times higher rate than pristine TiO
2, with the former fabricated using KMnO
4 to create trap sites to separate charges via bandgap impurity states
[63].
Depositing other catalysts has also been another approach. For example, a noble metal, such as Pt, Au, and Ru, can be deposited to increase the photocatalytic activity of TiO
2 towards the decomposition of organics and photocatalytic water splitting
[6][48][64][65][66]. This enhanced activity is likely due to improved charge separation as bulk electrons transfer to the metal and therefore to the surface of TiO
2 [6][64][65][66][67]. Though not all photogenerated electrons transfer from the titania to the metal, the enhanced separation of photogenerated charge carriers increases the electron lifetime in TiO
2. The separation is likely enhanced by the surface plasmon resonance (SPR) of the metal and the resulting increased localized EM field caused by the exposure of the metal to light. This induces charge carrier formation near the TiO
2 surface, with which carriers can easily reach surface sites and improve charge separation. Loading TiO
2 with gold instead of platinum also extends the absorption of TiO
2 to the visible range up to near-infrared
[6][68][69][70][71]. Using gold cores with TiO
2 shells also exhibits remarkable photocatalytic activity compared to TiO
2 due to enhanced separation of photogenerated charge carriers with the gold core serving as an electron trap
[72].
In addition to metals, metal oxides can also be deposited on TiO
2 to help improve charge separation and photocatalysis. For example, a Pd-NiO/TiO
2 catalyst has been prepared to improve photocatalytic CO
2 reduction. Due to the high electron density needed to drive this multielectron reduction, a p–n junction formed by introducing NiO to TiO
2 helps to drive hole transport to NiO, whereas the Pd forming a Schottky junction with TiO
2 facilitates semiconductor-to-metal electron transfer. These migrations towards NiO and Pd enhance the charge separation and result in high electron density around Pd, which can be used to transform CO
2 efficiently and selectively to CH
4 by reduction
[73].
The coupling of the semiconductor to TiO
2 can also be achieved by coupling it with another TiO
2 phase. As mentioned before, mixed-phase TiO
2 of anatase and rutile demonstrates improved photocatalytic performance than by using only pure anatase or rutile. This could be due to the formation of heterojunction when their valence and conduction band edges come into contact
[11][74]. Several models were proposed to explain this synergistic effect in mixed phases. First was the model proposed by Bickley et al. based on the positions of the CBs of anatase and rutile in relation to each other
[75]. Separation of charges then happens in anatase and trapping of an electron occurs in the rutile phase. In another “spatial charge-separation model,” Hurum and coworkers propose that the opposite is happening such that electrons are transferred from the rutile CB to a trapping site in anatase
[76].
Meanwhile, the “interfacial model” proposed by Nair et al. looks at the band bending at the interface between anatase and rutile. The electron transfer should occur from the anatase to the rutile upon UV illumination due to the CB energy of the anatase being more negative than that of the rutile. When the illumination is with λ > 380 nm, the rutile is activated, and its CB shifts upward due to accumulated photoinduced electrons enabling the electrons in the rutile to reach the CB of anatase
[46]. Thus, the “interfacial model” presents a directional movement of the electrons depending on whether the irradiation is ≤ or >380 nm and upon consideration of the interfacial band bending
[77]. One can expect that the interfacial nanostructure plays a role in the electron transfer between the components and therefore in the overall photocatalytic performance. Further discussion on the advantages of mixed-phase TiO
2 for photocatalysis can be found in the literature
[46].
Fe
2O
3 has also been combined with TiO
2 via photodeposition to enhance charge separation for contaminant decomposition and PEC water splitting. The achieved enhancement of more than 200% in complete mineralization kinetics was ascribed to the transfer of photoelectrons from the TiO
2 to the Fe
2O
3, which in turn favors the rate-determining step of oxygen reduction
[78]. Graphitic nanocarbon has also been added to TiO
2 nanomaterials to improve charge separation. By covering short single-wall carbon nanotubes (SWCNT) (~125 nm) around TiO
2 NPs (100 nm) using a hydration-condensation technique, longer lifetimes of photogenerated charge carrier and improved photocatalytic activity for the degradation of an aldehyde was achieved. This was better than nanographene and longer SWCNT hybrid systems. The shorter SWCNT provides greater interfacial contact with each TiO
2 NP, more electron transport channels, and more efficient shuttling of electrons from TiO
2 NP to SWCNT, delaying charge recombination. Improved SWCNT debundling with the short ones also affords these advantages to a larger portion of the composite
[79].
The semiconductor junction can also be quite complex, involving materials such as MXenes. For example, Biswal and coworkers designed a Ti
3C
2/N, S-TiO
2/g-C
3N
4 heterojunction to boost the spatial separation of charges and their transport in light of a photocatalytic water-splitting application
[79]. This heterostructure was produced by thermal annealing and ultrasonic-assisted impregnation for H
2 production that is up to ~4-fold higher than pristine S-doped titania. The dual heterojunction formed (a n–n heterojunction with a Schottky junction) likely not only enables effective charge carrier separation as CT channels
[79] but also reduces the band gap due to the adjustment of the energy bands. In(OH)
3-TiO
2 heterostructures were also formed for enhanced photocatalytic H
2 evolution. The band-gap tuning and improved charge separation resulted in up to a >15-fold increase in activity compared to commercial P25
[80].
4.2. Enhanced Light Harvesting
The morphology or nanostructure of TiO
2 also improves photocatalytic efficiency due to enhanced light harvesting. As mentioned above, nanoparticle aggregation (~500 nm) can improve light harvesting due to its high scattering effect, resulting in photon reabsorption. This increased visible-range absorption in turn increases the number of excited charges as seen in the increased current density. Such aggregates display unsmooth surfaces, resulting in better molecule adsorption in large surface areas and pore sizes
[81].
Nanotextured TiO2 substrate produced by nanomolding also displays efficient light harvesting. The hierarchical nanopattern of dual-scale nanoscale craters featuring smaller bumps couples both the longer and shorter wavelengths of light resulting in a light trapping effect for efficient light utilization and at a wide angular range
[82]. Similarly, cicada-wing-like structures were used as imprints to form nanohole structures of TiO
2 decorated with Ag NPs (10−25 nm) for methylene blue (MB) photocatalytic degradation. The structure did not only exhibit extended absorption to the visible range but also greater light absorption, likely due to the SPR effect from Ag and the nanotexture of TiO
2. This is based on the photocatalytic decomposition rate obtained for the Ag-TiO
2 nanotexture being 2.7 times higher than the nanotextured TiO
2 alone but more than 7 times higher than P25. This shows that even with just nanotextured TiO
2, improved photocatalytic performance can be seen. As discussed in the previous section, the Schottky barrier formed between Ag and TiO
2 could also improve the charge separation
[83]. Hollow particles of
TiO2 decorated with Au@Ag core–shell NPs also display enhanced light harvesting due to the combined strengths of the components of having a strong, broadened localized SPR, large specific surface area, and favorable light scattering properties
[84].
Orderly arrays of nanostructures, such as TNA, serve as effective light scattering layers according to the Mie scattering theory. The Mie scattering effect displayed by anodized TNA or NPs has been used in solar cells to harvest more sunlight and enhance charge conduction
[85]. Mie scattering is important in explaining particle size-dependent Raman enhancement observed with semiconductors
[86][87] and is brought about by the plasmon resonance at the surface of the sphere causing signal enhancement that depends on the size as one comes closer to the lowest transverse electric mode of NPs. In addition to the Mie effect, size quantization also affects the Raman intensity obtained on TiO
2 NPs
[88].
The surface-enhanced Raman (SER) effect on semiconductors has also been well-observed
[89][90][91][92][93], and the influence of nanostructuring on SER scattering, in particular on TiO
2, has been investigated. Whereas CT and chemical contribution can provide an enhancement factor (EF) of ~10
3 [94][95], EM enhancement is also afforded in TiO
2 nanostructures of a high aspect ratio, such as nanotubes and nanofibers
[96][97][98][99]. Han et al.
[98] showed concrete evidence of morphology-dependent EM enhancement using cyt b
5 heme as an indirectly-attached SER probe to reduce the chemical contribution to the Raman signals. Using EM field calculations, the particle’s aspect ratio was shown to increase the “hot spots” (regions of enhanced EM field) at the TiO
2-water interface
[98], improving the structure’s light-harvesting ability. Hence, other morphologies of higher anisotropy, such as TiO
2 nanotubes, were further studied, showing a similar morphology-dependent EM field enhancement
[96][99]. TNAs were shown to exhibit different Raman enhancements depending on the tube length, which fits the EM field calculation showing hot spots along the nanotube length
[96][99]. The TNA of high EM field enhancement was shown to perform better as a photocatalyst for visible-light-degradation of an azo dye pollutant immobilized on TNA. Interestingly, the TNA’s optical response (i.e., its EM field enhancement) correlates with the photocatalytic degradation rate occurring on it
[100].
Similar to other TiO
2 nanostructures (see above), incorporating other components to form nanocomposites with TiO
2 nanotubes can further improve not only the charge separation but also the light-harvesting ability. For example, S-doping or the addition of CdS NPs to TiO
2 nanotubes resulted in enhanced visible-light water splitting
[101][102][103]. Ultrafine Pt NPs were also added into TiO
2 nanotubes for the efficient photocatalytic formation of methane from carbon dioxide and water. The nanotubes allowed for a homogeneous distribution of Pt NPs, which accept electrons and become sites for reduction, thereby also allowing efficient separation of charges
[46]. Furthermore, even structures obtained from TNA somehow retain the light enhancement afforded by the 2D periodic arrangement of the nanotubes. For example, nitridation of the TNA resulting in a partially collapsed nanotube structure of TiN also shows wavelength-dependent EM field enhancement and corresponding light enhancement
[104].
The 2D periodic arrangement in TNAs enables them to behave as photonic crystals—with photonic lattices reflecting the light of certain wavelengths—bringing about localized EM field enhancement
[96][99][105][106][107][108]. Interestingly, this photonic crystal-like character has also been observed in inverse-opal (IO) structures, which also achieved SER EF of around 10
4 (though likely due to both chemical and EM contributions)
[96][99][105], and which can also be made from TiO
2. IO TiO
2 also shows promising performance as photocatalysts
[109][110][111][112], with their light harvesting extended to the visible range
[111][112] and their ability to generate slow photons
[112][113][114]. The slow photons have been shown to significantly increase the interaction of TiO
2 with light and can work synergistically to amplify the chemical enhancement in the catalyst
[110].
4.3. Black TiO2
The photocatalytic efficiency of TiO
2 nanomaterials can be improved by enhancing the separation of their charge carriers and improving their light harvesting and absorption properties (
Section 2.4.1 and
Section 2.4.2). Therefore, having a material that encompasses both is an ultimate surface-engineering achievement in this regard. Black TiO
2 ticks both requirements and reasonably has then become a hot topic in TiO
2 photocatalysis in the last decade or so.
Though previous studies already describe a similar material, as indicated in reviews
[115][116][117], all papers seem to point to the work of Chen et al.
[118] for introducing (and coining the term) “black TiO
2” to describe the partially hydrogenated titania nanocrystals which exhibit a reduced bandgap due to a disordered layer at the surface of its crystals. This material exhibited a redshifted absorption onset to near-infrared (compared to the starting TiO
2 nanomaterial), which was not surprising considering its visible color change. That is, due to the hydrogenation process, the crystals changed from white to black (hence the name). Consequently, this also results in a decreased bandgap of ~2.18 eV, making black TiO
2 a good catalyst for visible-light irradiation. Additionally, it also exhibits good stability, making it an ideal catalyst for use under continued irradiation. From calculations, it also presents localized photogenerated charge carriers, indicative of slowed-down recombination, which is beneficial for photocatalysis. This makes the work of Chen et al. the first reported use of black titania for photocatalytic purposes
[116].
From then on, many studies have been carried out to synthesize, characterize, and evaluate the photocatalytic performance of black TiO
2 nanomaterials. Different methods have been developed to reduce TiO
2 without the use of high pressure, as was conducted in the work of Chen et al.
[118]. These include (electro-)chemical reduction
[119][120], solvothermal hydrogenation
[121], thermal reduction
[122], reduction at the solid phase (reductant + heat)
[123], anodization (and annealing)
[124], ultrasonication
[125] plasma treatment
[126], gel combustion
[127], or a combination of these
[128]. Most of these strategies are similar to the synthesis of TiO
2 nanomaterials, with a reductant source/ reducing condition (either chemical, thermal, hydrogen, or reducing gases, such as hydrogen, nitrogen, and argon) added. Since black TiO
2 is formed by the reduction of TiO
2, it is also called “reduced TiO
2”
[129][130][131] or “hydrogenated TiO
2”
[132] and represented with the formula TiO
2−x, the −x indicating the formation of oxygen vacancies
[129][130].
What is interesting then is the concept of forming a novel material due to the introduction of surface defects, and yet, as this is also a TiO
2 nanomaterial, it can also exist in different morphologies and structural states, resulting in a plethora of black TiO
2 of various properties and photocatalytic performance.
Table 1 gives examples of these materials and their photocatalytic performance in terms of organic compound degradation and hydrogen generation, with the latter being an important solar-driven application of black TiO
2. Chen et al.
[117] give a comprehensive review of black TiO
2 nanomaterials, including their properties and examples of application, whereas Naldoni et al.
[116] give a good summary of the photocatalytic H
2 generation on black TiO
2. The readers are encouraged to take a look at these reviews.
Table 1. Examples of black (and colored) TiO2 nanomaterials and their photocatalytic performance in terms of degradation of organic compounds or hydrogen evolution. For table reprint/adaptation, please cite the original source: Querebillo, C.J. A Review on Nano Ti-Based Oxides for Dark and Photocatalysis: From Photoinduced Processes to Bioimplant Applications. Nanomaterials 2023, 13(6), 982; https://doi.org/10.3390/nano13060982. Copyright 2023 Querebillo, C.J.
Some examples included in
Table 1 are of different color naming (termed “colored TiO
2”), such as green, grey, and blue TiO
2 [116][119][132][134][135]. This is based on the understanding that the visual colors exhibited by TiO
2 are brought about by intrinsic defects, such as due to the presence of Ti
3+ and/or oxygen vacancies
[116][119][124][132][135] or by doping with impurities
[116][126]. Such defects create extra electronic states in the TiO
2 bandgap, i.e., intraband gap states, which alters the optoelectronic properties of TiO
2 [116]. Whether this is also the case for the color of black TiO
2 is still a controversy. While some reports claim that the formation of these color-inducing intrinsic defects in TiO
2 results from the hydrogenation
[129], with the color depending on the extent of reduction and reducing condition
[132], others propose that the black color is due to the disordered surface
[118][126]. The disordered surface is caused by hydrogen and allows hydrogen to swiftly navigate around and induce electronic structural changes
[136]. Midgap band states are formed because of the changes in the structure
[127] brought about by the excessive lattice disorder. They can form an extended energy state by overlapping with the edge of the conduction band and also possibly combining with the valence band
[126].
An effort to further unravel the relationship between the defect nature and photocatalytic activity of reduced TiO
2 was performed by Will et al.
[132] by considering that the intrinsic defects created on the surface are pertinent to the photocatalytic process and the location of the defect depends on the structural state and reducing conditions. They found that the introduction of Ti
3+ at the surface results in a surface with substoichiometry, which activates the surface for photocatalysis. However, too long hydrogenation or too much Ti
3+ is detrimental to its activity, as these provide additional recombination sites or prevent efficient interfacial CT. Surface roughness and strain were also not important for the activation of photocatalysis.
The photocatalytic activity of black TiO
2 nanomaterial can be further enhanced by forming appropriate heterojunctions with other materials
[132], a similar strategy used with TiO
2 nanomaterials. Further, amorphous black TiO
2 can also be synthesized and used for photocatalysis
[125], which is important for applications such as for bone implants. Black TiO
2 was shown to exhibit biocompatibility
[137], regenerative properties
[138], photothermal properties
[137][138][139], and microbicidal action
[137][139][140][141], among others. As such, black TiO
2 shows promise for cancer treatment
[138][139], as a bone implant coating, and for disinfection purposes
[137][139][140][141]. Similar to TiO
2, the photo(electro)catalytic disinfection with black TiO
2 nanomaterials is also claimed to occur due to ROS, in particular, superoxide and/or hydroxyl radicals
[140][141]. Nevertheless, the biosafety of black TiO
2 needs to be further studied to intensify its application in the biomedical field.
This entry is adapted from the peer-reviewed paper 10.3390/nano13060982