On the Origin of Enhanced Photocatalytic Activity of Copper-Modified Titania in the Oxidative Reaction Systems

Abstract

Modification of titania with copper is a promising way to enhance the photocatalytic performance of TiO₂. The enhancement means the significant retardation of charge carriers’ recombination ratio and the introduction of visible light activity. This review focuses on two main ways of performance enhancement by copper species—i.e., originated from plasmonic properties of zero-valent copper (plasmonic photocatalysis) and heterojunctions between semiconductors (titania and copper oxides). The photocatalytic performance of copper-modified titania is discussed for oxidative reaction systems due to their importance for prospective applications in environmental purification. The review consists of the correlation between copper species and corresponding variants of photocatalytic mechanisms including novel systems of cascade heterojunctions. The problem of stability of copper species on titania, and the methods of its improvement are also discussed as important factors for future applications. As a new trend in the preparation of copper-modified titania photocatalyst, the role of particle morphology (faceted particles, core-shell structures) is also described. Finally, in the conclusion section, perspectives, challenges and recommendations for future research on copper-modified titania are formulated.

1. Introduction

Titanium dioxide (TiO₂, titania) has played a crucial role in the development of semiconductor photocatalysis over the last 40 years [1]. The applicative potential of TiO₂-photocatalysis is presently focused on the areas such as environmental remediation (water treatment [2,3,4] and air purification [5]), renewable energy processes—water splitting for hydrogen production [6], conversion of CO₂ to hydrocarbons [7], solar cells [8]—and self-cleaning surfaces [9]. TiO₂ is characterized by those properties that are indispensable to fulfill the requirements of efficient, stable, and green photocatalytic material (long term stability, chemical inertness, corrosion resistance, non-toxicity) [1].

Despite the above-mentioned benefits, the application of titania is still limited to regions with a high intensity of solar radiation due to its wide bandgap (ca. 3.0 to 3.2 eV). Therefore, there is a necessity to incorporate visible light absorption with TiO₂. Various strategies of titania performance improvement towards visible light responses have been proposed in recent years such as doping, modification, semiconductor coupling, and dye sensitization [10]. As a consequence of such modifications, TiO2 showed an absorption band in the visible light region and, in most cases, also photocatalytic activity under visible light irradiation for different reaction systems.

The photocatalytic activity of titania is also limited by the recombination of the photogenerated electron-hole pairs [11] as typical for all semiconducting materials. Titania exhibits only weak bandgap emission upon the recombination of conduction band electrons with valence band holes, and the irradiative recombination involving trap states is optically allowed [1,12]. Recombination is generally caused by impurities, defects or other factors, which introduce bulk or surface imperfections into the crystal and, depending on titania properties, it may occur either on the surface or in the bulk (usually a decrease in particle size results in an increase in surface recombination) [13,14]. The incorporation of species capable of promoting charge separation may reduce this phenomenon, and thus result in improvement of overall quantum efficiency of photocatalytic system. Promoted charge separation has been reported for titania modified with various modifiers, e.g., ions [14,15], noble metals [16,17], other semiconductors (heterojunction coupling) [18,19], and doped with metal ions [14].

Considering the above-mentioned limitations (vis inactivity and charge carriers’ recombination), and similar ways of their removal (surface modification, doping, heterojunction), it is thought that effective materials based on TiO₂, which meet those requirements are highly designable. It must be pointed out that modification of titania towards visible light photocatalytic activity could significantly influence photocatalytic performance under UV irradiation and in some cases a decrease in photocatalytic activity has been observed since modifiers/dopants could also work as recombination centers for electrons and holes, therefore a proper selection procedure for titania modification should consider both limitation issues.

Among different metallic candidates for TiO₂ modification, copper (Cu) is a very promising material. Cu-based nanocatalysts have significant applicability in nanotechnology including catalytic organic transformations, electrocatalysis, and photocatalysis [20,21,22]. In comparison with other noble metals (gold, platinum and silver) recognized as very efficient co-catalysts of titania, copper—as a consequence of its abundance in the Earth’s crust—is an inexpensive material, 100 times and 6000 times cheaper than silver and gold, respectively [21]. Furthermore, owing to the location in the same group of the periodic table as gold and silver, copper has similar properties due to its electronic configuration and the face centered cubic (FCC) structure of the atom’s location. Therefore, one can expect the comparable potential to improve the photocatalytic activity of TiO₂, simultaneously with the higher possibility of successful application. Copper can exist in a wide range of accessible oxidation states: Cu⁰, Cu^I, Cu^II and Cu^III. In this connection, the active copper species in TiO₂ photocatalytic system are copper oxides (Cu₂O, CuO) and metallic copper. Cu may also co-exist in both forms (the oxide and zero-valent) and this is the most probable case when titania is modified with metallic copper under anaerobic conditions, and thus-obtained photocatalyst is subsequently kept under aerobic conditions (surface oxidation of metallic copper deposits). The variety of copper forms can provide difficulty in understanding of their role in considered reaction systems. Therefore, a statement that the nature of Cu species in TiO₂ photocatalysis is still not clearly understood may be made. There is also some possibility of copper ions doping into TiO₂ crystal lattice. The radius of Cu²⁺ is 0.087 nm, which is larger than the 0.0745-nm radius of Ti⁴⁺. Moreover, a huge difference in valence state suggests that Cu²⁺ should not replace Ti⁴⁺ to enable displacement doping at the crystal lattice site [23]. There is only the possibility of incorporating Cu²⁺ ions into interstitial positions in the lattice [24,25,26,27].

Various papers on copper-modified titania have been already published, including comprehensive reviews [22,28] about copper species in photocatalysis. However, these reviews are mainly focused on the photocatalytic reduction reactions, i.e., hydrogen evolution and/or CO₂ to methane conversion. Of course, the Cu/TiO₂ system is most known for the mentioned reaction systems—and the number of papers in this field is the highest—but the application of Cu/TiO₂ for the photocatalytic oxidation of organic compounds is also very promising both under ultraviolet (UV) and visible/solar irradiation, and the role of Cu species in this system needs clarification. Therefore, in the present work, an overview of papers has been carried out to compare different mechanisms resulting from an existence of various forms of copper combined with titania in relation to efficiencies of oxidative reaction systems.

2. Photocatalytic Systems Based on Cu/TiO₂

2.1. Presence of Copper Ions in the TiO₂ Suspension

The first papers about Cu/TiO₂ systems were mainly connected with photodeposition of copper on titania surface as the efficient method for Cu removal from water environments [29,30,31,32]. Cu^II ions were reduced to metallic copper with the participation of photogenerated electrons (e⁻) and holes (h⁺) according to the following Reactions (1)–(3) [29]:

TiO₂ + hν → e⁻ + h⁺

(1)

Cu²⁺ + 2e⁻ → Cu

(2)

H₂O + 2h⁺ → 0.5O₂ + 2H⁺

(3)

In opposition to the above results of Bard’s group [29], Hermann et al. suggested that Cu²⁺ ions in TiO₂ suspension were photoreduced to Cu⁺ ions and that metallic copper could not be obtained [33]. However, Jacobs et al. showed that photodeposition of zero-valent Cu particles was possible and was preceded by photodeposition of Cu₂O particles [32].

Another type of study focused on the influence of the presence of copper ions on the efficiency of photocatalytic oxidation of organic compounds. It was found that addition of dissolved copper ions to TiO₂ reaction system improved significantly the rate of photocatalytic oxidation [34,35,36,37,38,39]. For example, Bard et al. observed [29] that Cu²⁺ ions retarded the electron-hole recombination by trapping of photogenerated electrons (Equation (2)). The reduced forms, in turn, could prevent the recombination by hole trapping (Equation (4)).

Cu + 2h⁺ → Cu²⁺

(4)

Generally, the small concentration of Cu²⁺ ions prevents charge recombination. In the case of large amounts of Cu²⁺ ions, their influence on the reaction rate can be detrimental considering that Reactions (2) and (4) are short-circuiting. In 1985, Okamoto et al. showed that a high concentration of Cu²⁺ (50 mM) inhibited the efficiency of phenol photooxidation (the first study using copper for photocatalytic oxidation of organic compounds) [34]. The detrimental effect of a high amount of Cu²⁺ ions was also explained by absorbance of UV/vis radiation by copper species [40]. Precipitation of dissolved copper as a hydroxide could also decrease the rate of photocatalytic oxidation by reflecting UV irradiation through increased solution turbidity (shielding effect) [39]. Brezova et al. observed an inhibition effect of dissolved Cu²⁺ ions even at low concentrations (>1 mM) in the phenol oxidation system. The key point is that the observed photodeposition of metallic copper and Cu₂O on titania may significantly influence the process of charge carriers and radical intermediate generation and recombination and finally, along with Equation (2), result in the inhibition of photodegradation [40]. Butler and Davis observed negligible adsorption of copper (and other metals) on titania surface and suggested that mainly dissolved copper (and other metals) increased the reaction rate via a homogeneous pathway rather than surface reactions on copper-modified titania [39,40]. They proposed a mechanism involving formation of a ternary reactive complex between copper, organic compound or its oxidation intermediate and oxygen-containing species such as O₂, H₂O₂. Bideau et al. analyzed TiO₂ system in the presence of Cu²⁺ ions for the oxidation of three organic compounds: formic acid, acetic acid and propionic acid, and monocarboxylate complexes of Cu²⁺ were identified as important species in the reaction kinetics [35,36,37,41]. The observed increase in the reaction rate in the mentioned studies was explained in terms of the effects of Cu²⁺/Cu⁺ redox couple, which inhibited the electron-hole recombination and the inner sphere mechanism of Cu²⁺ ions with the organic compounds, forming organo-metallic intermediates.

2.2. Copper Species Deposited on TiO₂ Surface

2.2.1. Metallic Copper—Plasmonic Photocatalysis

Copper (similar to gold and silver) nanoparticles (Cu NPs)–titania couple can be a very promising photocatalytic system for both UV and visible light-induced reactions. Cu NPs as surface modifiers of TiO₂ can inhibit electron-hole recombination under UV irradiation working as an electron sink [42,43], because Fermi energy level of metallic Cu lies below the conduction band of TiO₂, and therefore photogenerated electrons in TiO₂ can be easily transferred to Cu NPs. Other important advantage of Cu NPs is that they can activate TiO₂ towards visible light due to localized surface plasmon resonance (LSPR) of Cu, thus being so-called “plasmonic photocatalysts” [44,45,46,47,48,49]. The LSPR is excited when light interacts with the free electrons of a metallic nanostructure, which results in the collective excitations (oscillations) that lead to significant improvement of the local electromagnetic fields surrounding the nanoparticles [50]. Spherical metallic nanoparticles, such as gold, silver and copper, are sufficiently small (in comparison with the wavelength of the light) to be resonant with the light coming from all directions and indicate coloration as well as a local electromagnetic field enhancement on the particle surface even without the condition of total internal reflection [44]. The existence of characteristic surface plasmon resonance band absorbing light in the visible region results in visible-light activity of noble metal-modified titania, probably due to an efficient transfer of the photoexcited electrons from metal particles to the conduction band of TiO₂ (energy transfer and plasmonic heating have been also proposed as possible reasons of vis response [51]). This would result in electron-deficiency in metal and electron-richness in TiO₂ and therefore, the direct photocatalytic oxidation occurs on the metal surface rather than on TiO₂ surfaces as shown in Figure 1 [52] or indirect oxidation proceeds by reactive oxygen species (initiated by superoxide anion radical formation).

Based on the unique properties of LSPR, many studies focused mainly on Au and Ag as plasmonic metals, which improve photocatalytic properties of TiO₂ [53]. However, very few studies have been reported about photocatalytic performance of Cu NPs in connection with titania [47,48,49]. The main problem with utilizing Cu NPs as a “plasmonic sensitizer” is the known fact that zero-valent copper is easily oxidized and lose plasmon resonance properties gradually under ambient conditions [54,55,56]. For example, (i) although, titania modification by strong radiolytic reduction of Cu²⁺ or photodeposition under anaerobic conditions resulted in formation of zero-valent copper, Cu⁰ was subsequently oxidized under ambient conditions forming CuO/TiO₂ and Cu/Cu₂O/Cu/TiO₂, respectively [57,58,59] and (ii) even Cu NPs immersed in water were oxidized by dissolved oxygen [60]. The reported main solutions of this problem consider the necessity of preventing copper oxidation by maintaining Cu in anoxic environments [55,61,62,63] or protecting its surface with chemical corrosion inhibitors [64], polymeric layers [47,55], and by oxide encapsulation [65]. The optimal Cu NPs based plasmonic photocatalysts should work without necessity of addition of surface-obscuring chemical stabilizers and would be storable indefinitely under ambient conditions [49]. Most of these studies were focused on Cu NPs alone but not on Cu NPs-TiO₂, where it is necessary to consider interactions between Cu NPs and TiO₂ during preparation procedure [66].

Most studies about Cu NPs-TiO₂ were connected with unprotected copper under reductive, oxygen free conditions. Zhang et al. prepared TiO₂ nanotube arrays with copper NPs by pulsed electrochemical deposition method [48]. They reported visible light activity of this material for hydrogen evolution. Similarly, Kum et al. investigated the properties of the visible light active Cu/TiO₂ photocatalyst for hydrogen production [67]. This material showed high stability and photocatalytic activity in the evacuated chamber. However, it is difficult to distinguish whether the photocatalyst was activated by excitation of copper LSPR or band gap heterojunction with copper oxides since photocatalysis experiments were driven by broadband visible light irradiation.

In order to apply the Cu NPs-TiO₂ system for an oxidative system, the persistence of metallic copper to oxidative environment is necessary. Yamaguchi et al. prepared plasmonic Cu NPs deposited on TiO₂ electrode and protected by polyvinyl alcohol, which resulted in efficient photocurrent response due to LSPR [47]. In the latest research, De Sario et al. studied titania aerogels supporting ca. 5-nm Cu NPs [49]. They found that plasmonic properties of Cu NPs were preserved by offering an extended interfacial contact with reduced TiO₂ support (characteristic design issue for aerogels). An arrangement of Cu NPs at high-surface area TiO₂ (with multiple Cu-TiO₂ contacts per Cu nanoparticle) allowed the obtaining of a Cu/TiO₂ system with good stability and a high content of Cu NPs. Figure 2 shows DRS spectra of Cu/TiO₂[aerogel] (with characteristic peaks for copper LSPR at ca. 770 nm) before and after photoelectrochemical oxidation of methanol confirming high stability of this material (almost the same plasmonic properties of Cu/TiO₂).

2.2.2. Copper Oxides—Heterojunction Systems with TiO₂

The Concept of Heterojunction between Copper Oxides and Titania

It is well known that coupling TiO₂ with other semiconductors with different redox energy levels can lead to an increase in photocatalytic efficiency by the enhancement of the charge carrier separation process and thus an increase in the lifetime of the charge carriers. Moreover, the proper selection of coupled semiconductors can also activate the heterojunction system towards a visible light response. To create the efficient heterojunction system, both semiconductors must possess different energy levels from their corresponding conduction and valence bands. In such a configuration, several advantages can be obtained: (a) an improvement of charge carriers’ separation; (b) an increase in the lifetime of the charge carrier; and (c) an enhancement of the interfacial charge transfer efficiency to adsorbed substrates [68,69].

In the case of an efficient interparticle electron transfer between the semiconductor and TiO₂, the conduction band of TiO₂ must be more anodic than the corresponding band of the sensitizer. Under visible light irradiation, only the semiconductor-sensitizer is excited and the electrons photoexited to its conduction band are injected into the TiO₂ conduction band. If the valence band of the sensitizer is more cathodic than the valence band of TiO₂, the hole generated in the semiconductor remains there (enable to migrate to TiO₂). These thermodynamic conditions favor the phenomenon of electron injections, as shown in Figure 3a. When the system of coupled semiconductors works under UV–vis irradiation, both semiconductors are excited. For heterojunction system described above, two origins of electrons in the conduction band (CB) of titania are considered: (i) injected into TiO₂ (the same as under vis excitation) from coupled semiconductor; and (ii) photoexcited electrons from titania valence band under UV irradiation, resulting in a high concentration of electrons in CB of TiO₂. Whereas, holes left in the valence band (VB) of TiO₂ may migrate to the VB of coupled semiconductor and influence a high concentration of holes in the couple semiconductor/electrolyte interface (Figure 3b) [70].

Cu₂O and CuO are p-type semiconductors with band gap energies of 2.1 eV and 1.7 eV, respectively. The band positions of copper oxides in relation to TiO₂ are shown in Figure 4a.

Depending on the band positions of two semiconductors forming heterojunction, three different types of heterojunction could be distinguished: I, II and III (Figure 4b). Type II, which provides the optimum band positions for efficient charge carrier separation, is considered the most probable for the Cu_xO/TiO₂ system. Photoexcited electrons are transferred from CB(B) to CB(A) and this transfer can occur directly between semiconductors due to favorable energetics of the relative positions of CBs, or due to band bending at the interface inducing and internal electric field. Whereas, holes are transferred simultaneously from VB(A) to VB(B) and as a result photogenerated electrons and holes are separated from each other, reducing the recombination probability and increasing the lifetimes of the charge carriers [71].

Both copper oxides, especially Cu₂O, are very promising semiconductors for photocatalytic hydrogen production [72,73,74,75]. The main limitation of their application results from not good photostability, which is very important issue for oxidative photocatalytic systems. Generally, Cu₂O possesses higher photocatalytic activity than CuO for degradation of organic compounds [75,76]. Although CuO has much smaller band gap than Cu₂O, and thus is able to absorb more vis photons, the positions of CB and VB for CuO are insufficient to catalyze the production of hydroxyl and superoxide radicals, which are primary initiators for the photocatalytic oxidation of organic compounds [75]. Deng et al. found that combined CuO/Cu₂O nanostructures can be efficient photocatalysts for the photodegradation of organic compounds and possess better resistance against photocorrosion [76]. They proposed that the co-existence of CuO could inhibit the photocorrosion of Cu₂O, and furthermore the CuO/Cu₂O system can be more photocatalytically efficient than single CuO or Cu₂O.

CuO-TiO₂ Heterojunction

Taking into consideration above-mentioned issues, the heterojunction between p-type semiconductors–copper oxides and n-type one–titania could be a very promising way to improve the photocatalytic activity of titania under UV or/and visible light irradiation and to obtain more active and stable photocatalytic material than copper oxides alone. In the middle of eighties Okamoto et al. prepared Cu-deposited TiO₂ by UV-irradiation of deaerated suspension of slightly reduced (in hydrogen) anatase powder in the presence of CuSO₄ [34]. They assumed that only metallic copper was obtained on titania surface (but without experimental evidence). It was found that Cu-deposited TiO₂ was significantly more active in phenol photocatalytic oxidation than single TiO₂ in the presence of Cu²⁺ ions. However, the useless of this photocatalyst for wastewater treatment was suggested because of dissolution of deposited Cu into the reaction solution transforming again to Cu⁺ and Cu²⁺ ions. This research expressed two main problems with the discussion on Cu/TiO₂: careful characterization of copper forms connected with titania, and the stability of prepared photocatalyst. To understand the mechanism of improvement of photocatalytic activity and to analyze perspectives of application, thorough research considering above issues is necessary.

The first research including aspects of different forms of copper in connection with titania and photooxidation activity was performed by Song et al. in 1999 [77]. They investigated the effect of oxidation state of copper loaded on titania surface on photocatalytic oxidation reaction. Cu was photodeposited on TiO₂ surface in the presence of CuSO₄ and ethanol as a hole scavenger. Moreover, to obtain Cu/TiO₂ photocatalysts with different oxidation states of surface-loaded copper, samples were annealed at 200 °C and 300 °C. They reported that non-annealed samples contained only metallic copper, whereas annealing at 200 °C and 300 °C resulted in copper oxidation to mainly Cu₂O and CuO, respectively. It was found that metallic copper deposited on titania improved the photocatalytic activity of TiO₂ for photooxidation of 1,4-dichlorobenzene. In contrast, the use of oxidized form of copper on titania resulted in a significant decrease in photocatalytic activity and it was proposed that copper oxide clusters on TiO₂ were not effective in the transfer of photoexited electrons [77].

Another report, which confirmed detrimental effect of CuO presence, was done by Chiang et al. for photocatalytic oxidation of cyanides in the CuO/TiO₂ system [78]. CuO/TiO₂ was prepared by photodeposition method and subsequent thermal treatment in air at 110 °C for various copper contents (0.05–10 at. % (atomic percentage) Cu). Only slight improvement of photocatalytic activity was observed for 0.1 at. % Cu, due to suggested electron trapping by CuO. It was proposed that a decrease in photocatalytic activity with a further increase in CuO content was caused by reduced photon absorption since well dispersed nanosized CuO particles covered the surface of TiO₂ (shielding effect). Moreover, it was suggested that higher concentration of CuO could promote recombination of photogenerated holes with the trapped electrons resulting in a decrease in available holes for redox reactions. Chiang et al. also examined photocatalytic activity in cyanide oxidation system under the presence of dissolved Cu²⁺ ions. The presence of copper ions was detrimental for the rate of reaction. This decrease could be explained in terms of the competition reaction of copper(I) cyanide complexes for the photogenerated hydroxyl group [78].

Shun-Xin et al. proposed another preparation method of Cu/TiO₂, i.e., impregnation of titania precursor in copper salt during sol-gel synthesis [79]. They suggested the co-existence of both copper oxides on titania surface and copper ions (Cu²⁺ and Cu⁺) in the lattice of titania (doping). Authors proposed that Cu²⁺ could trap the excited electron in CB of titania, inhibiting electron-hole pair recombination, whereas Cu⁺ could transfer electron to oxygen adsorbed on the surface of photocatalyst accelerating interfacial electron transfer [79].

Arana et al. studied methyl tert-butyl ether (MTBE) photocatalytic oxidation on CuO-TiO₂ under UV irradiation [80]. They observed a significant improvement of photocatalytic activity of copper-modified titania in comparison with pure TiO₂. Considering the reduction potential for Cu²⁺/Cu⁺ (+0.17 V vs. NHE (normal hydrogen electrode)), CuO deposits on TiO₂ surface could react with the photogenerated electrons through Reaction (5), and Cu⁺ ions could be re-oxidized to Cu²⁺ by oxygen, H₂O₂ or other oxidizing species present in the Medium (6):

Cu²⁺ + e⁻_CB → Cu⁺

(5)

Cu⁺ + (O₂, H₂O₂, other oxidants) → Cu²⁺ + e⁻

(6)

Therefore, the Reaction (5) of Cu²⁺ ions with the photogenerated electrons should slow down electron-hole recombination resulting in activity enhancement [80].

The mechanism for the Cu/TiO₂ system under visible light irradiation was proposed by Irie et al. for 2-propanol oxidation [81,82]. They prepared Cu(II)-grafted TiO₂ photocatalyst by impregnation method using CuCl₂·2H₂O as the source of copper. Authors suggested that Cu²⁺ ions existed in the form of amorphous CuO clusters. They proposed that visible light irradiation initiated interfacial charge transfer (IFCT): electrons in the VB of TiO₂ were directly transferred to Cu²⁺ and Cu⁺, and the holes created in the VB of TiO₂ decomposed organic compounds (Figure 5) [82]. It was proposed that for efficient IFCT, Cu²⁺ ions had to be atomically isolated on the TiO₂ surface. Authors suggested that produced Cu⁺ ions could reduce O₂ by a multielectron reduction, being re-oxidized to Cu²⁺. This multi-step reduction process could be initiated by a two-electron reduction of oxygen to peroxide (Equation (7)), and the following four-electron reductions (Equations (8) and (9)).

2Cu⁺ + O₂ + 2H⁺ → 2Cu²⁺ + H₂O₂

(7)

3Cu⁺ + O₂ + 4H⁺ → 2Cu²⁺ + Cu³⁺ + 2H₂O

(8)

4Cu⁺ + O₂ + 4H⁺ → 4Cu²⁺ + 2H₂O

(9)

It was concluded that Cu²⁺ ion in the amorphous form affords the smooth reversibility between Cu²⁺ and Cu⁺ forms [82]. The possibility of this mechanism was also confirmed by this group for other material, i.e., Fe(III)-grafted TiO₂ [83]. Similar mechanism was proposed for CuO/WO₃ system by Arai et al. [84]. Photoexcited electrons in WO₃ were transferred to CuO and Cu²⁺ was reduced to Cu⁺, and reduced surface was re-oxidized (Cu⁺ to Cu²⁺) by oxygen.

The properties of photocatalytic material prepared by Irie et al. [82] were optimized in the field of crystallinity, the interfacial junction between TiO₂ and CuO nanoclusters, and the amount of CuO nanoclusters. The synthesized CuO-TiO₂ nanocomposites exhibited efficient interfacial charge transfer (IFCT) for decomposition of volatile organic compounds and strong anti-pathogenic effects under indoor conditions. The mechanism of those dual type properties is presented in Figure 6. It was proposed that Cu⁺ species were very efficient for enhancing the antibacterial and anti-viral properties of TiO₂ photocatalysts (even under dark conditions), i.e., holes in the valence band (VB) of TiO₂, generated under irradiation with visible light, in combination with Cu⁺ species, could attack the outer membrane, proteins, and nucleic acid (DNA and RNA) of viruses and bacteria, resulting in their death and inactivation [85].

To understand the mechanism of coupled photocatalysts, it is necessary to know the exact positions of the valence band (VB) and conductance band (CB) of semiconductors. The ambiguity over the positions of the conduction and valence bands in CuO (and Cu₂O) results in different explanations for the mechanistic aspects of heterojunction systems. For example, Li et al. prepared CuO/TiO₂ (rutile) nanocomposites by impregnation and subsequent calcination in air [86]. They tested photocatalytic properties of this material using methylene blue as a model compound under UV/vis irradiation. The authors proposed that CuO/TiO₂ system could be classified as type I of heterojunction, as presented in the inset of Figure 7 (different CB and VB positions for CuO than presented in Figure 4a). In this system, since the band edges of TiO₂ bracketed those of CuO, coupling between CuO and TiO₂ could lead to decreased photocatalytic activity due to transfer of both electrons and holes simultaneously from TiO₂ to CuO resulting in their recombination in CuO. However, when the copper content was very low (0.1 wt %) significant improvement of photocatalytic activity was observed in comparison to unmodified titania (Figure 7). Therefore, it was proposed that copper existed in this sample as highly dispersed CuO clusters and substitional Cu²⁺ (Ti-O-Cu linkages), and UV illumination resulted in trapping of electrons in TiO₂ lattice. Photogenerated electrons were localized at Ti sites in the bulk of titania lattice and holes at interfacial sites of TiO₂-CuO (by EPR spectroscopy). However, at higher copper loadings, no photogenerated electrons were observed at titania lattice trapping sites. Slamet et al. explained this phenomenon as a “shading effect” [87] (also known as “shielding effect” or “inner filter effect” and observed for various titania surface modifications, e.g., with Au NPs [88], and Pt NPs [89]): colored CuO absorbed light and reduced the photoexcitation capacity of TiO₂ (competition about photon absorption). Li et al. concluded that efficient electron transfer from the CB of TiO₂ to CuO led to the absence of Ti³⁺ (trapped electrons) in CuO-TiO₂ at high CuO loadings. They formulated a hypothesis that the form of CuO (clusters or nanocrystallites) had a significant influence on the type of charge transfer kinetics between TiO₂ and CuO [86].

Moniz et al. prepared a CuO/TiO₂ (P25) composite using microwave co-precipitation technique [90]. It was found that 5% CuO/P25 exhibited near 1.6 times higher efficiency than pure P25 in mineralization of a model organic pollutant (2,4-dichlorophenoxyacetic acid (herbicide)) under UV/Vis. To analyze the mechanism of photocatalytic activity, the authors determined the band edge positions of CuO and TiO₂. They experimentally determined those positions by the interpretation of impedance (Mott-Schottky) plots. The difference in conduction band energies of the two materials suggested that electron transfer from TiO₂ to CuO was feasible, given the 0.34 V offset which could provide sufficient driving force for charge separation (Figure 8). It was also evidenced by the fact that pure CuO was inactive for photo-decomposition, and thus photocatalytic oxidation by CuO did not occur [90]. The concept of this junction was not connected with the increase of visible light absorption of P25 but only with more efficient charge separation. Photogenerated electrons were transferred from TiO₂ to the conductions band of CuO, followed by transfer of photoexcited holes from TiO₂ to the surface for the expected oxidation reactions.

Siah et al. prepared CuO photodeposited on TiO₂ [91]. Under UV irradiation, the photocatalytic activity for 2,4-dichlorophenoxyacetic acid decomposition increased by a maximum factor of 4.3 compared with unmodified TiO₂, when TiO₂ was loaded with 0.75 mol % CuO. In contrast, under visible light and solar simulator irradiation, the optimum loading of CuO was much lower (0.1 mol %), and enhancements of 22.5 and 2.4 times were observed. In the case of samples loaded with higher amounts of CuO (1–5 mol %), CuO clusters can mask the surface of titania, which prevents the light source from reaching the active sites. CuO clusters in CuO-high loaded TiO₂ may act also as recombination centers. Therefore, in addition to the increased charge separation and improved visible light absorption, the masking effect will also be considered in designing efficient heterojunctions in the copper oxides–titania system [91].

Another strategy to obtain efficient visible light active photocatalyst, based on titania–copper(II) oxide, was proposed for self-doped TiO₂ (Ti³⁺) grafted with amorphous CuO [92]. Cu-modified self-doped TiO₂ was obtained by thermal treatment of mixture of titanium oxides (Ti₂O₃ and TiO₂ mixture) in air, and their subsequent impregnation with copper salt solution. As it was reported in earlier Irie’s works [81,82], in the case of titania modified with amorphous nanoclusters of Cu²⁺ oxide, electrons in the valence band of TiO₂ could be directly excited to CuO nanoclusters, which could serve as an efficient oxygen reduction site through the multi-electron process. Significant improvement of visible light activity was achieved for CuO-modified-self-doped titania (CuO/Ti³⁺-TiO₂) in comparison with CuO/TiO₂ in the oxidation of gaseous 2-propanol. Under visible light irradiation, induced electrons on an isolated Ti³⁺ band could be transferred to the surface CuO nanoclusters efficiently, in addition to the direct charge transfer from the titania VB to the CuO nanoclusters, as shown in Figure 9. Therefore, for CuO/Ti³⁺-TiO₂, under visible light irradiation, there were holes in the VB decomposed 2-propanol, while photoinduced electrons were consumed (via oxygen reduction) on CuO nanoclusters. It was proposed that amorphous Cu(II) oxide nanoclusters grafted on titania surface suppressed the recombination of electron—hole pairs (at the isolated Ti³⁺ band), and acted as a co-catalyst for efficient oxygen reduction to consume the photoinduced electrons. Therefore, the photocatalytic activity of CuO/Ti³⁺-TiO₂ under both UV and visible light irradiation was very high and stable [92].

Cu₂O-TiO₂ Heterojunction

For the first time, a Cu₂O/TiO₂ heterojunction system without the presence of other forms of copper was prepared by electrochemical deposition of Cu₂O on Ti foil by Siripala et al. as a thin film heterojunction photocathode [93]. They observed a photoresponce demonstrating efficient light-induced charge carriers’ separation at Cu₂O-TiO₂ junction. Moreover, it was proposed that TiO₂ film successfully protected the Cu₂O layer against photocorrosion. In the next study, Li et al. prepared composites of Cu₂O-TiO₂ (P25) by electrochemical method for UV-decomposition of dye brilliant red [94]. Prepared material was more active than pure P25, but unfortunately the mechanism was not discussed.

Bessekhouad et al. studied photocatalytic efficiency of Cu₂O/TiO₂, prepared by direct mixing of both semiconductors [70]. The activity of this heterojunction system was checked by using model organic pollutant Orange II under both UV and visible light irradiation. It was assumed that Brownian motion sufficed to permit charge transfers between the particles of the two solids. They observed that the amount of Cu₂O in the heterojunction played an important role, since vis photocatalytic efficiency increased with an increase in Cu₂O amount reaching saturation at ca. 30% (Figure 10a). It was proposed that under visible light irradiation, electrons from Cu₂O could be injected into the CB of TiO₂, and at titania surface react with dissolved oxygen molecules and induce a formation of oxygen peroxide radicals (O₂^•−). Similarly, in the case of the UV system, an increase in content of Cu₂O resulted in enhanced efficiency, but resultant activities at high content of Cu₂O were only slightly higher than that of pure TiO₂, whereas for low content of Cu₂O a significant deactivation of TiO₂ was noticed (Figure 10b). This fact was attributed to the increase of charge trapping mechanism contributions when both semiconductors are excited. According to the authors, the CB of Cu₂O was more negative than the CB of TiO₂ favoring the electron transfer from Cu₂O to TiO₂. Other conditions, such as the kinetics of reactions occurring at each surface, should be also considered. For example, the holes generated in Cu₂O had to react faster to induce an efficient charge separation [70]. Moreover, it should be reminded that lack of holes’ consumption (oxidation reaction by holes) could result in photocorrosion of the sensitizer (here Cu₂O) [69].

Huang et al. performed similar investigations on Cu₂O/TiO₂ system under UV and visible light irradiation for decomposition of Orange II [95]. They used titania P25, Cu(CH₃COO)₂·H₂O as a copper source and alcohol-aqueous based chemical precipitation method to prepare heterojunction photocatalysts, in which copper was present in only one form-Cu₂O (2–3-nm particle size). Similar to Bessekhouad work, it was found, that an increase in content of Cu₂O resulted in higher photocatalytic activity (70%-Cu₂O content with highest activity). They reported 6 and 27 times higher photocatalytic activity for Cu₂O/TiO₂ system than that for pure P25 for UV–Vis and visible light irradiation, respectively. The weak point of this and Bessekhouad’s research under visible light irradiation is the fact that vis-response should not be tested for dyes due to possibility of semiconductor sensitization [96,97]. Although Huang et al. justified that the excited electrons could not migrate from Orange II to Cu₂O, since the reduction potential of Acid Orange II was lower (−1.25 V) than CB of Cu₂O (−1.5 V), the possibility of electron migration to CB of titania could not be omitted. According to Figure 11, under UV light irradiation, both Cu₂O and TiO₂ could be excited (process (1) and (3) corresponding to Equations (10) and (11), respectively), the generated electrons in Cu₂O and holes in TiO₂ could migrate to CB of TiO₂ (process (2)) and VB of Cu2O (process (4)), respectively. This transfer process is thermodynamic favorable since both CB and VB of Cu₂O lie above that of TiO₂ (heterojunction type II), which results in prolongation of the lifetime of excited electrons and holes, inducing higher quantum efficiency. Meanwhile, the generated electrons react with dissolved oxygen molecules and produce O₂^•− (Equations (12) and (13)) to decompose organic pollutants (Equation (14)), while the holes could induce some oxidation process directly (Equation (15)). Under visible light irradiation, process (1) can also occur due to the narrow band gap of Cu₂O (2.0 eV), and thus Orange II still can be photocatalytically decomposed by the redox reactions (Equations (11)–(14)), where Equation (12) is possible due to the transfer of electrons from Cu₂O (process (2)):

Cu₂O + hν → Cu₂O(e⁻) + Cu₂O(h⁺)

(10)

TiO₂ + hν → TiO₂(e⁻) + TiO₂(h⁺)

(11)

TiO₂(e⁻) + O₂ → TiO₂ + O₂^•−

(12)

Cu₂O(e⁻) + O₂ → Cu₂O + O₂^•−

(13)

OII^*_ads + O₂^•− → products

(14)

OII^*_ads + TiO₂(h⁺) → products

(15)

Similar conclusions were proposed by Zhang et al., who prepared TiO₂ film covered by a Cu₂O microgrid by the microsphere lithography method [98]. The underlying TiO₂ film was composed of nanosized particles covered by microgrids of Cu₂O, which were composed of particles smaller than 20 nm. By comparison of three types of thin films—TiO₂, Cu₂O and TiO₂-Cu₂O for methylene blue degradation—it was found that under both UV and visible light irradiation the heterojunction system was the most active. The authors suggested the same mechanism of heterojunction as that explained above (Figure 11).

A modified explanation of the enhanced photocatalytic activity of Cu₂O/TiO₂ heterojunction system considering the role of Ti³⁺ was proposed by Xiong et al. [99]. TiO₂ and Cu₂O were prepared separately and final coupled photocatalyst was obtained by suspension mixing. It was reported that Cu₂O/TiO₂ photocatalyst had much better vis activity in degradation of brilliant red X-3B and photocatalytic hydrogen evolution than TiO₂ and Cu₂O alone. It was proposed that photogenerated electrons from Cu₂O were captured by Ti⁴⁺ ions in TiO₂ and thus being reduced to Ti³⁺ ions, as shown in Figure 12. Whereas, left holes in the valence band of Cu₂O formed “hole centers” hampering the charge carriers’ recombination. Therefore, trapped electrons (in Ti³⁺ ions with prolonged lifetime) could be transferred to the interface between the composite and solution, resulting in retarded recombination between electrons in TiO² (Ti³⁺) and holes in Cu₂O [99].

Wang et al. prepared Cu₂O-loaded TiO₂ nanotube arrays (NTAs) by electrochemical anodization of Ti foil and subsequent Cu₂O deposition by an ultrasonication-assisted sequential chemical bath deposition (S-CBD) method [100]. They check both visible light photocatalytic (Rhodamine B as model pollutant) and photoelectrocatalytic activities of Cu₂O/TiO₂ NTAs. The photocatalysts with small content of Cu₂O exhibited the largest photocurrent and photoconversion efficiency under visible light irradiation, as well as the highest visible light photocatalytic degradation rate of RhB. In particular, when 0.5 V bias potential was applied, Cu₂O/TiO₂ NTA photoelectrodes were found to possess superior photoelectrocatalytic efficiency, due to a synergistic effect of electricity and visible light irradiation [100].

However, in the case of heterojunctions for photocatalysts obtained according to standard methods an improved charge separation was observed, some limitations were also suggested, due to the incompact contact and small interface between two semiconductors. Therefore, the preparation of interconnected heterostructures with a large interface between two semiconductors was proposed for improvement of the charge transfer efficiency. For example, Cu₂O@TiO₂ core-shell photocatalyst was prepared by in situ hydrolysis and crystallization method (Figure 13), and tested under solar radiation for 4-nitrophenol degradation [101]. It was found that Cu₂O@TiO₂ had absorption properties similar to pure Cu₂O: strong visible light absorption in the range of 400–600 nm (Figure 14a), and was the most active among other coupled photocatalysts, e.g., prepared by physical mixing Cu₂O/TiO_2(PM) and by chemical method denoted to Cu₂O/TiO₂ (Figure 14b). Moreover, weaker photoluminescence signals for Cu₂O@TiO₂ sample than for pure Cu₂O suggested retardation of charge carriers’ recombination, probably because introduction of TiO₂ shell could reduce the surface oxygen vacancies and defect in Cu₂O core [101].

Another strategy for obtaining efficient copper oxide–titania coupled photocatalyst is based on consideration of the morphology of copper oxide and titania. For example, Liu et al. synthesized Cu₂O nanospheres decorated with TiO₂ nanoislands (Cu₂O-NS/TiO₂-NI) by a facile hydrolyzation reaction followed by a solvent-thermal process. [102]. It was found that Cu₂O-NS/TiO₂-NI demonstrated superior photocatalytic performance under visible light illumination for methyl orange degradation, E. coli bacteria disinfection, and also a better stability during the photocatalysis process from their specific structure. The design of a partial coverage of Cu₂O nanosphere with TiO₂ nano-islands enabled the reaction between photo-generated holes and water to produce hydroxyl radicals or directly with organic pollutants/microorganisms in water. Therefore, the accumulation of holes on the underlying Cu₂O films prevented photocorrosion, and subsequently made this photocatalyst stable during the photocatalysis process (Figure 15). It was shown that part of Ti⁴⁺ was reduced to Ti³⁺ in Cu₂O-NS/TiO₂-NI, suggesting that under visible light irradiation photo-generated electrons in Cu₂O were transferred to TiO₂ and trapped there. After shutting off the irradiation, these trapped electrons could be gradually released from TiO₂ to react with O₂ producing reactive radicals, which imparted the photocatalytic memory effect to Cu₂O-NS/TiO₂-NI photocatalyst (Figure 15) [102].

It is well known that controlled morphology by application of exposed facets (faceted semiconductors) significantly enhances photocatalytic performance, due to excellent crystallinity and low content of defects, e.g., octahedral and decahedral anatase particles [103,104]. Therefore, it is not surprising that similar concept was applied for copper oxide–titania heterojunction systems. For example, Liu et al. prepared Cu₂O@TiO₂ photocatalysts using Cu₂O with exposed facets [105]. First, Cu₂O nanocrystals with different types of morphology: cubes, cuboctahedra and octahedra were synthesized, and then Cu₂O@TiO₂ core-shell polyhedral nanostructures were prepared by hydrothermal reaction. It was proven then the morphology of Cu₂O cores was well preserved during hydrothermal process (Figure 16A). Well-crystallized anatase nanoparticles created a uniform and rough layer covering the smooth surface of Cu₂O core. The photocatalytic activity of Cu₂O@TiO₂ in the visible light was checked in relation to methylene blue and 4-nitrophenol. The activity of all coupled samples was higher than that of single Cu₂O and TiO₂, and increased with the following order regarding Cu₂O morphology: cubes < cuboctahedra < octahedra. The surface photovoltage spectroscopy (SPS) was used to analyze photoinduced carrier separation and charge transfer behavior, where the magnitude of SPS response peak depended on the amount of net charge accumulated on the material surface. It was shown that SPS response correlated with the type of facet configuration, and Cu₂O@TiO₂ octahedra had the strongest SPS response, Cu₂O@TiO₂ cuboctahedra—medium, whereas Cu₂O@TiO₂ cubes—the lowest (Figure 16B), showing the same tendency as the photocatalytic activity. Moreover, it was found that Cu₂O@TiO₂ core-shell photocatalysts had different band offset values (Figure 16C), demonstrating a clear facet-dependent activity. The photocatalytic performance of Cu₂O@TiO₂ polyhedra was different with that of pure Cu₂O polyhedra without TiO₂ shells, which could be attributed to the different driving forces for the charge carrier separation [105].

Mixed Copper Species Deposited on Titania Surface

Frequently, during preparation of copper-modified TiO₂, a mixture of copper species is obtained [58,59], and such systems are named as “cascade heterojunction systems.” For example, Helaili et al. prepared Cu₂O/TiO₂, Cu/Cu₂O/TiO₂ and Cu/Cu₂O/CuO/TiO₂ heterojunctions using chemical methods, and examined their potential application as photocatalysts for oxidative reactions (Orange II as a model dye pollutant) under visible light irradiation [106]. The obtained results suggested the following rules, which could govern the electrons transfer mechanism: (a) metallic Cu improved the photocatalytic activity of single semiconductor by the formation of apparent Ohmic junction enhancing the charges transfer kinetics; (b) electromotive forces developed from semiconductor/semiconductor heterojunction suppressed the effect of metallic Cu due to the formation of an apparent Schottky type junction; (c) diffusion potential developed from multiple steps electrons transfer had more impact for improving the efficiency than that developed from one step transition; and (d) heterojunction cascade obtained from monobloc of photosensitizers was more efficient than that obtained from direct mixture of the semiconductors [106]. Xing et al. prepared Cu/Cu₂O/TiO₂ photocatalyst as an efficient heterojunction system for terephthalic acid degradation under UV/Vis irradiation [107]. According to the mechanism presented in Figure 17, under UV/vis irradiation electrons and holes are photogenerated on both Cu₂O and TiO₂. Because the Fermi energy level of Cu is lower than that of _p-type Cu₂O, electrons from the CB of Cu₂O migrate to Cu until the two Fermi levels are aligned. Holes in the VB of Cu₂O recombine with electrons in the CB of TiO₂. Therefore, holes in the VB of TiO₂ are efficiently separated from electrons, and engage in hydroxyl radicals’ formation. In the case of visible light irradiation, only Cu₂O is photoexcited. Electrons in the CB of Cu₂O are trapped by highly conductive Cu, whereas migration of holes from the VB of Cu₂O to VB of TiO₂ is impossible, and thus, the negligible amount of hydroxyl radicals can be detected. This transmission pathway (under UV) is analogous to that of the Z-scheme system instead of a type II heterojunction system. It means that Cu/Cu₂O/TiO₂ system with very intensive charge separation is much more efficient in the case of UV/vis system than for visible light system, where Z-scheme model dominates [107].

Analogically, it is thought that the morphology of titania should be also important issue for the efficiency of copper species–titania heterojunction system. Janczarek et al. prepared copper-modified titania based on decahedral anatase particles (DAP) with eight equivalent (101) facets and two (001) facets [58]. Copper species were photodeposited on DAP surface. It was estimated that Cu₂O was a predominant form of copper (82.0%), whereas other copper species existed in minority, CuO (12.8%) and Cu(0) (5.2%). Copper species formed small nanoclusters of ca. 2 nm, which were uniformly distributed on all facets of DAP. This material Cu/Cu₂O/CuO/TiO₂(DAP) was especially active in UV/vis system for acetic acid oxidation but very low improvement of activity for 2-propanol oxidation was observed under visible light irradiation, confirming mechanistic issues described in the previously mentioned studies [58].

Qiu et al. prepared Cu_xO-TiO₂ nanocomposites dedicated to indoor environments for degradation of volatile organic compounds and pathogens [108]. The structure of copper species deposited on titania surface was nanocluster mixture of Cu₂O and CuO. The balance between Cu(I) and Cu(II) states in Cu_xO was critical to achieving efficient VOC decomposition and antipathogenic activity. It was found that the optimum Cu(I)/Cu(II) ratio in Cu_xO nanoclusters was 1.3. This material had also good antiviral and antibacterial activity under dark conditions. The ratio of Cu(I)/Cu(II) did not markedly change after long-term visible-light irradiation, indicating that the multielectron reduction reaction was catalytic in air with a turnover number greater than 22 [108]. Luna et al. studied Cu₂O-CuO-TiO₂(P25) semiconductors for gallic acid oxidation under UV/vis LEDs irradiation [109]. By using UV light, CuO addition to TiO₂ did not enhance the photoactivity and acted like an impurity or a defect at the surface, favoring the fast recombination of charge carriers, or like an electron fast scavenger. In the case of visible light system (Figure 18), CuO addition evidenced the injection of charge carriers from Cu₂O to TiO₂ and CuO, as suggested by TRMC study. Gallic acid degradation proceeded by a sequence of reactions involving the formation of intermediates such as, maleic, fumaric, oxalic and formic acids, before reaching its mineralization [109].

3. Conclusions

The design of an efficient and stable photocatalyst for oxidative reaction systems with high activity under UV and visible light is still a challenge. Among various metals considered as titania modifiers, copper is one of the most promising because of its abundancy, low cost and similar properties to expensive noble metals such as gold and silver. The enhancement of TiO₂ photocatalytic activity with copper modification can be realized by two main mechanistic concepts: preparation of plasmonic photocatalyst with metallic copper on titania surface and formation of heterojunctions between titania and copper oxides. Direct doping of titania lattice by copper ions is rather difficult because of different sizes of Cu²⁺ and Ti⁴⁺. The chemical stability of copper species is a very important issue for copper-modified TiO₂. This issue should be always considered in design thinking about copper-modified titania, primarily for metallic copper. In this review, the successful attempts to obtain stable metallic Cu were presented, and this breakthrough research may cause an increase in the number of research in the copper-focused, visible light-oriented plasmonic photocatalysis.

This review also highlights various heterojunction concepts according to different copper oxides–titania configurations. Owing to the authors’ knowledge there is still no comprehensive research paper considering all main variants of copper oxides–titania heterojunctions to perform a comparable study. To analyze the mechanism of heterojunction, it is necessary to calculate exact positions of VB and CB of semiconductors. The problem is that these values for copper oxides can significantly differ in various papers, resulting in change of the possible type of heterojunction. Therefore, an accurate photoelectrochemical characterization of each component of heterojunction system is required. Among copper oxides in relation to heterojunction with titania, Cu₂O has the promising application potential to wastewater treatment including its very good antipathogenic properties, even better that metallic copper [108]. For oxidative reaction systems, cascade heterojunction with different forms of copper is also a perspective option. Additional presence of metallic copper in such heterojunction system is beneficial under UV irradiation, but is detrimental in the case of visible light irradiation. The presence of Ti³⁺ ions can also positively influence the efficiency of heterojunction system. Moreover, the content of copper oxides (Cu₂O/CuO) in cascade heterojunction system is a very significant parameter influencing photocatalytic activity. The new research areas are opening with the development of methods, which allow to obtain copper oxides and titania with specific particle morphology. The influence of different types of crystal facets of Cu₂O on the efficiency of Cu₂O/TiO₂ system has been confirmed and is still under debate. Furthermore, the core-shell configurations for copper oxide–titania heterojunctions can increase a photocatalytic efficiency and stability of whole photocatalytic system. Additionally, the post-illumination catalytic memory phenomenon, observed in Cu₂O(nanospheres)/TiO₂(nanoislands) can be important extension for photocatalytic properties under dark conditions.