Update on Interfacial Charge Transfer (IFTC) Processes on Films Inactivating Viruses/Bacteria under Visible Light: Mechanistic Considerations and Critical Issues

This review presents an update describing binary and ternary semiconductors involving interfacial charge transfer (IFCT) in composites made up by TiO2, CuO, Ag2O and Fe2O3 used in microbial disinfection (bacteria and viruses). The disinfection mechanism, kinetics and generation of reactive oxygen species (ROS) in solution under solar/visible light are discussed. The surface properties of the photocatalysts and their active catalytic sites are described in detail. Pathogenic biofilm inactivation by photocatalytic thin films is addressed since biofilms are the most dangerous agents of spreading pathogens into the environment.


Introduction
This update is designed primarily for students and researchers in the area of photocatalysis, describing the design, synthesis, evaluation and characterization of innovative photocatalysts allowing environmental disinfection. The advantage of supported photoactive materials is that they preclude the use of chlorinated compounds in water bodies to inactivate bacteria/viruses. Chlorination generates dangerous/carcinogenic chloro-species due to the interaction of Cl radical reacting with organic compounds in water under solar irradiation.
Semiconductor and inert oxides/double oxides present a mild alternative approach to disinfect water bodies. No comprehensive mechanism for the degradation of organic matter and pathogens by photocatalysis is available at the present time. There is a need for more work on stable, kinetically fast photocatalysts inactivating virus/bacteria. Recent reviews reporting environmental disinfection methods have been published recently [1][2][3][4][5][6]. This is a topic of timely interest and will motivate more research on the understanding of viral resistance to catalytic/photocatalytic treatment.
Most studies available on the photocatalytic inactivation of viruses rely on TiO 2 , a low-cost, inert, non-toxic semiconductor that is stable within a wide pH range. However, the performance of TiO 2 is limited by its slow inactivation kinetics induced under solar light, resulting in long treatment times. This precludes a wide application of this approach. Doped TiO 2 promoters have been reported to enhance the virus inactivation kinetics by TiO 2 at a modest scale [5]. Highly porous TiO 2 increasing the adsorption of dispersed virus/bacteria is a way to increase the virus disinfection kinetics. Highly porous films such as TiO 2 /Ag, TiO 2 /Cu and TiO 2 /Cu-apatite are not available at the present time [3][4][5].
Viruses constitute a group of heterogeneous organisms. They range in size from 0.01 to 0.3 microns and are much smaller compared to bacteria. They do not present independent metabolism and are fed by the host. They contain either DNA or RNA as genetic materials, but not both. The nucleic materials are surrounded by a protein and are coated to protect radical R leads to the formation of further ROS in solution. This is shown schematically in Figure 1b. ride (LPS) bilayer and leads to lipid radicals, as shown in Figure 1a. Within longer times, the oxidative radical R leads to the formation of further ROS in solution. This is shown schematically in Figure 1b.
The E. coli envelope consists of lipopolysaccharides (LPS) and phosphatidyl-ethanolamine (PE) bilayers, and these bilayers photodegrade during TiO2-mediated photocatalysis. Within longer times, the most resistant layer of the envelope, the peptidoglycan (PGN), gets damaged, leading to several peroxidation products. Depending on the type of bacteria used, the degradation products consist of aldehydes, ketones, carboxylic acids and residual organic oxidized intermediates. Full mineralization of the bacteria leads to CO2, water and inorganic anions such as NO3 − /SO4 = [13,14]. Light irradiation to inactivate viruses has revealed that adenoviruses are highly resistant to irradiation by UV [15]. The United States Environmental Protection Agency (USEPA) has increased the UV fluence requirements for 4-log10 removal of viruses from 40 to 186 mJ/cm 2 . This latter dose was shown to be necessary to completely deactivate the virus. This implies a higher amount of energy/costs necessary to inactivate this kind of pathogens. More efficient photocatalysts are required at the present time. Adenovirus is resistant to conventional treatment since its genome is mainly DNA and not RNA. This enables the virus to utilize the DNA repair enzymes present in the host cells to repair its genome damage, increasing its resistance to disinfection. Metals can undergo specific re- The E. coli envelope consists of lipopolysaccharides (LPS) and phosphatidyl-ethanolamine (PE) bilayers, and these bilayers photodegrade during TiO 2 -mediated photocatalysis. Within longer times, the most resistant layer of the envelope, the peptidoglycan (PGN), gets damaged, leading to several peroxidation products. Depending on the type of bacteria used, the degradation products consist of aldehydes, ketones, carboxylic acids and residual organic oxidized intermediates. Full mineralization of the bacteria leads to CO 2 , water and inorganic anions such as NO 3 − /SO 4 = [13,14]. Light irradiation to inactivate viruses has revealed that adenoviruses are highly resistant to irradiation by UV [15]. The United States Environmental Protection Agency (USEPA) has increased the UV fluence requirements for 4-log 10 removal of viruses from 40 to 186 mJ/cm 2 . This latter dose was shown to be necessary to completely deactivate the virus. This implies a higher amount of energy/costs necessary to inactivate this kind of pathogens. More efficient photocatalysts are required at the present time. Adenovirus is resistant to conventional treatment since its genome is mainly DNA and not RNA. This enables the virus to utilize the DNA repair enzymes present in the host cells to repair its genome damage, increasing its resistance to disinfection. Metals can undergo specific redox covalent reactions with the topmost functional groups of the virus or cells' outer layers. This involves redox reactions between the metal/ions and generates organometallic toxic species, leading to cell inactivation in the dark or under light [16][17][18][19].
Singlet oxygen has been reported to inactivate MS2 by the oxidation of amino acid residues [3][4][5]20]. Binary TiO 2 -SiO 2 -mediated photocatalysts have shown improved virus inactivation ability when compared to bare TiO 2 due to the finer dispersion of the TiO 2 nanoparticles (NPs) reached on the SiO 2 . Hydrogen peroxide was found to be an effective additive in the dark for the reduction of viruses on stainless steel and, when used in conjunction with UV, led to a faster inactivation compared to H 2 O 2 applied alone. For this reason, combined hydrogen peroxide-UV treatments to reduce viruses from many foods have been used during recent years.

Enterococcal Viruses
The enteric group of viruses are not able to replicate in the environment outside of their hosts. Visible light-irradiated TiO 2 was seen to produce ROS in solution, leading to the inactivation of rotavirus and astrovirus. Human noroviruses are the most common cause of viral gastroenteritis. Fluorinated TiO 2 films have been shown to more effectively inactivate human norovirus under visible light compared to TiO 2 by itself. One hour of irradiation with fluorescent light at 10 mW/cm 2 reduced norovirus approximately 3-4log 10 . When no light was applied, no detectable reduction was observed. The norovirus destruction was due to ROS and was accelerated by the presence of F-radicals leaching out of the catalyst surface [6]. Different viruses have been found in treated sewage leading to infections, despite the information reported by municipal treating stations that they are complying with the legislation on the standard microbiological water treatment. This revealed insufficient municipal treatment of wastewater viruses and enteric bacteria [15]. A more stringent evaluation of biologically treated water is needed nowadays to ensure the quality of drinking water and ensure that it is free of viruses.

Poliovirus
Poliovirus was effectively inactivated by titanium dioxide photocatalysis. The rates were more rapid than for the inactivation of coliform bacteria. However, the photocatalytic disinfection of effluents using titanium dioxide under sunlight is severely limited due to the relatively slow inactivation kinetics. Polio virus 1 was 99.9% killed in secondary waste effluent after 30 min of irradiation with 40-W black light in phosphate buffer Dulbecco's Phosphate Buffered Saline (DPBs) TiO 2 P25 dispersions [2].

Influenza
The influenza virus has been shown to be inactivated by TiO 2 photocatalysis. First, the protein present was deactivated, and subsequently, RNA was attacked. The data suggest that the photocatalytic activity destroys the viral binding proteins. The degradation of the proteins was due to ROS, mainly involving • OH and O 2 • -produced by the TiO 2 photocatalysis. The virus inactivation depends on the UV-A intensity and irradiation time. Titanium dioxide photocatalysis under UV light can inactivate 4-log 10 influenza virus within a short time at extremely low doses of UV-A (0.01 mW/cm 2 ) [21,22].

Adenovirus
Adenoviruses are found widely in contaminated food or sewage water. UV-C treatment has been used at a laboratory scale and induced a higher reduction of virus than UV-A and UV-B. Adenoviruses' inactivation in the water supply chain is, in some instances, carried out using monochloramine with the generation of undesirable DBPs. Adenoviruses' contamination of water systems allow adenovirus to cycle through the environment. It is also transmitted from person to person, as in the case of the rotavirus. This virus spreads by respiratory/oral pathways by contact with the host cells of the receptor. After contact with the host cell, adenovirus produces the proteins required for further genetic replication. Adenoviruses proceed when significant damage takes place on their capsid structure. In contrast to MS2, adenovirus proteins were inactivated within acceptable times during the TiO 2 -mediated photocatalytic process. The adenovirus undergoes damage to its genomes by UV254 light. However, even when damaged, is still capable of infecting cells. Viruses use the host cell enzymes to repair their damaged DNA and then replicate, allowing for further infection of a variety of host cells.

Parameters Controlling the Photocatalytic Virus Inactivation
The sections below briefly describe the solution parameters controlling pathogen inactivation in solution. Most of the available scientific literature reports studies on TiO 2 deactivation of bacteria, but not viruses. The photocatalytic inactivation of viruses follows the methodology and photochemical set-ups used to inactivate bacteria. The experimental use is adjusted for the features of viruses.

Effect of the Catalyst Concentration
A higher catalyst concentration accelerates the virus inactivation kinetics, but this process is limited by the optical density of the solution. The limit is set by the light penetration into the reactor volume [1][2][3][4][5]. Increasing the concentration of TiO 2 or TiO 2 /Ag enhanced ROS production and led to an enhanced virus inactivation. The NP concentration mediating virus inactivation should be much higher than the photons/sec/cm 2 reaching the photochemical reactor. Saturation with the incoming photons of the photocatalyst NPs by the solution is another limiting parameter to consider during virus inactivation processes. Silver NPs were reported to enhance the TiO 2 photocatalytic inactivation kinetics of viruses. Silver also has a high affinity for sulfur (cysteine) and carboxyl groups present in the topmost virus bilayer. The latter functional groups bind silver [23,24].

Effect of Light Intensity
A higher light intensity leads to higher bacterial photocatalytic inactivation kinetics, but in certain limits. It increased linearly with the applied light dose when lower light intensities were used. However, the reaction rate only increased with the square root of the intensity at intermediate intensities. At higher light intensities, the rates increase only marginally or not at all. It was observed that the inactivation rates became constant after a certain threshold [25]. The reason for this is that higher light intensities lead to a faster recombination of the photogenerated charges (electrons and holes) in semiconductors such as TiO 2 [1][2][3][4][5]. The range and dose of the applied light have a controlling effect on the bacterial inactivation kinetics. Figure 2 shows the range of the light applied in photocatalytic processes. TiO2 coatings and metal /oxide composites have been reported to inactivate a wide variety of microorganisms in small-or large-size reactors [1][2][3][4][5]. The variation of the light dose within a day had important effects on the pathogen inactivation kinetics, as the  TiO 2 coatings and metal /oxide composites have been reported to inactivate a wide variety of microorganisms in small-or large-size reactors [1][2][3][4][5]. The variation of the light dose within a day had important effects on the pathogen inactivation kinetics, as the wavelength of the incident light changes during daily solar irradiation as a function of the time. UV-A and UV-B have been widely reported to inhibit cellular activity in the presence of TiO 2 or doped TiO 2 , and the observed inactivation kinetics is a function of (a) the type of virus and/or bacteria colonies to be inactivated and (b) the concentration of pathogens. A bovine rotavirus was inactivated within 3 h and a 3log 10 concentration reduction was observed under solar light irradiation. Poliovirus was inactivated by applying a solar dose of 85 mW/cm 2 at 25 • C < 6 h. E. coli was disinfected under identical conditions within shorter times.

Contact between the Virus and the TiO 2 Surface
A higher surface area of a TiO 2 NP photocatalyst has been reported to lead to faster bacterial inactivation kinetics compared to samples prepared at higher temperatures and presenting lower specific surface areas (SSAs) [26]. A closer contact of the TiO 2 NPs with bacteria/viruses has been reported to accelerate the pathogen inactivation kinetics [27]. The binding of the TiO 2 NPs to bacterial cells involves electrostatic, van der Waals forces, pH, agglomeration of the catalyst and the NPs' surface potential.

Photocatalytic Mechanism Leading to Virus/Bacterial Inactivation
TiO 2 photocatalysis processes damage protein, leading to viral capsids and releasing DNA, involving protein and genome degradation [5,6]. The inactivation sequence of a virus involves the following steps: (a) modification of the protein sequence cross-linking, (b) disruption of the protein conformation, (c) disruption of the protein aggregated size and (d) disruption of the transfer of the viral genome to the host cells. The surface area of the TiO 2 and its size, crystallographic structure and porosity control the amount of ROS produced in the solution leading to pathogen inactivation. The salts, cations and anions found in the solution interfere with the contact between the catalyst and the pathogen, hindering the bacterial/virus inactivation. Impurities and additives can also interfere and partially inhibit the adsorption of the pathogen onto the TiO 2 NPs.
Organic compound sensitizers are found in natural water bodies. Solar sensitizers transmit their energy by contact/diffusion to waterborne viruses [3][4][5][6][7][8]. These organic sensitizers adsorb solar irradiation and interact with the O 2 dissolved at normal temperature and pressure in water (8 mg/L). Sensitizers are activated by visible light when reacting with the dissolved O 2 present in the ground state as a triplet state. Triplet-triplet energy transfer occurs between S* and 3 O 2 or by a second deactivation channel electron transfer from the S* to oxygen. Alternatively, the electron can transfer from the S* to the superoxide radical anion O 2 − in the solution at biological pH values of 5.7-8.0. The sensitizer (S) can either transfer the energy, leading to excited O 2 * by reaction (1), or react subsequently by charge transfer with dissolved O 2 , as shown in reaction (2): The excited O* species and the radicals O 2 •− in Equations (1) and (2) and the lifetime found is of τ is 0.1 ms. The diffusion of the HO 2 • in solution is estimated as x = 2000 Å, taking the average reaction rate k for HO 2 • radicals as~10 6 M −1 ·s −1 . ROS leading to virus inactivation considers (a) that mass transfer does not play a role in the virus ROS transfer process and (b) that no activation energy is required for ROS radical reactions with viruses. The quenching of singlet oxygen 1 O 2 by NaN 3 has been used frequently to identify 1 O 2 and quantify its presence in solution. However, this approach presents a drawback: azides also react with the • OH radical at a rate close to its reaction with 1 O 2 . Singlet oxygen 1 O 2 has been reported with a lifetime of 3.6 ms and a diffusion motility of~75 nm in solution. This length is above the observed diffusion length for the • OH radical motility in solution of~70 Å. The reason for this is that the singlet oxygen is a species in the gas phase and the OH radical is a species in solution. 1 O 2 -mediated processes lead to E. coli, adenovirus and MS2 and many other pathogens' inactivation [27,28]. Singlet oxygen has also been reported to hinder MS2 genome replication and oxidize virus protein.
The OH radical presents an oxidation potential • OH/OH − 1.9 eV NHE [5] and is high enough to lead to inactivation of the MS2 virus and oxidize capsid proteins. However, the quantification of the intermediate ROS leading to virus inactivation has not been fully reported until now.
Bacteria and viruses can sorb on dissolved organic matter (DOM) in solutions and water bodies. This increases the probability of encounters between ROS and the DOM particles adsorbed on the surface of bacteria/viruses. DOM-virus interactions have been reported for capsid proteins due to cation bridges, electrostatic, steric, hydrophobic and carboxylate effects.

IFCT Ag/TiO 2 -Mediated Bacterial/Virus Inactivation
In recent years, photocatalytic heterostructures have been reported, presenting increased absorption in the visible region. This is attained by decorating/doping TiO 2 and ZnO semiconductors with metals, oxides and double oxides absorbing in the visible region. In the latter case, IFCT processes take place and (a) increase the separation of charge carriers in the semiconductor, (b) suppress the recombination rate of photoinduced electron-hole pairs, thus improving photocatalytic efficiency, (c) induce synergistic effects induced by the components in the heterostructure and, (d) in some cases, increase photostability of the composite [29].
Metallic oxides with oxygen vacancies such as W-oxides are of growing interest in the field of visible light-driven photocatalysis as they exhibit a wide absorption tail in the Near Infrared Region (NIR). This is due to oxygen defects. The combination of TiO 2 with poly-oxometalla (tes leads to a hybrid photocatalyst absorbing in the UV and visible region [30]. Ag is the oldest antimicrobial agent and has been known for 2000 years. Its antibacterial features have been reported in a series of reviews [5][6][7]. Silver is known to be antimicrobial, biocompatible and non-toxic to humans at levels < 0.1 mg/L. The virus inactivation kinetics mediated by TiO 2 /Ag is enhanced by the photogenerated charges in the TiO 2 surface increasing the Ag disinfection activity due to the presence of ROS radicals in solution. Higher Ag contents decrease the amount of light reaching the TiO 2 surface and reduce the generation of electrons, precluding the ability to inactivate pathogens. The interfacial charge transfer (IFCT) induced by TiO 2 nanoparticles (TiO 2 /Ag NPs) in solution under visible light is shown below in Figure 3.  Ag surface plasmon resonance (SPR) sites build up in Ag clusters of TiO2/Ag surfaces and this favors light absorption. TiO2/Ag increases the generation of ROS radicals such as HO2 •− and • OH radicals and 1 O2, leading to the degradation of viruses such as adenovirus, rotavirus and astrovirus. TiO2/Ag has been reported to increase the sorption/adsorption of pathogens compared to bare TiO2 in the dark and under light. The TiO2/Ag photocatalytic action is commonly rationalized in terms of an enhanced charge separation at the surface of the TiO2 by the Ag dopant hindering the e-/h+ charge recombination.
Variation of the of the rutile/anatase ratio in the catalyst make-up led to a variation of the kinetics of MS2 virus inactivation. Ag photocatalysis leads to the formation of ROS species [23,24]. Ag reacts almost instantaneously with air (O2) and atmospheric water vapor, leading to AgOH. Only 2-3 atomic layers of AgOH (0.4-0.6 nm thick) are found on NPs' surfaces [31]. AgOH decomposes spontaneously to Ag2O.
Ag2O is thermodynamically stable in the pH 6-8 range, the physiological pH range of bacteria. Visible light irradiation activates Ag2O, generating electrons and holes: Ag2O + light → Ag2Ocbe − + Ag2Ovbh + (6) and the photogenerated build-up of O2 •− , • OH and HO2 • radicals as outlined next: Ag surface plasmon resonance (SPR) sites build up in Ag clusters of TiO 2 /Ag surfaces and this favors light absorption. TiO 2 /Ag increases the generation of ROS radicals such as HO 2 •− and • OH radicals and 1 O 2 , leading to the degradation of viruses such as adenovirus, rotavirus and astrovirus. TiO 2 /Ag has been reported to increase the sorption/adsorption of pathogens compared to bare TiO 2 in the dark and under light. The TiO 2 /Ag photocatalytic action is commonly rationalized in terms of an enhanced charge separation at the surface of the TiO 2 by the Ag dopant hindering the e-/h+ charge recombination.
Variation of the of the rutile/anatase ratio in the catalyst make-up led to a variation of the kinetics of MS2 virus inactivation. Ag photocatalysis leads to the formation of ROS species [23,24]. Ag reacts almost instantaneously with air (O 2 ) and atmospheric water vapor, leading to AgOH. Only 2-3 atomic layers of AgOH (0.4-0.6 nm thick) are found on NPs' surfaces [31]. AgOH decomposes spontaneously to Ag 2 O.
Ag 2 O is thermodynamically stable in the pH 6-8 range, the physiological pH range of bacteria. Visible light irradiation activates Ag 2 O, generating electrons and holes: and the photogenerated build-up of O 2 •− , • OH and HO 2 • radicals as outlined next: Light excitation on catalysts leads to faster pathogen inactivation compared to runs in the dark. This is due to the additional release of • OH radicals as suggested next in the reaction (11) AgOH h + + H 2 O ads → OH ads + H + The Langmuir-Hinshelwood (L-H) kinetic model is frequently employed to describe bacterial inactivation kinetics [6,7]. This model is applied to process the data obtained during virus inactivation mediated by TiO 2 and metal-TiO 2 photocatalysts [5][6][7][8]. In the L-H model, the reactants are adsorbed and considered in equilibrium with the catalysis surface. This model assumes (a) that the system is in a steady state, (b) that the product formation is the rate limiting step, (c) that the products desorb quickly after the reaction and, finally, (d) that one of the reactants is present in a large excess. Viruses are made up of complex structures, presenting features that differ from chemical compounds, and the L-H model does not take into account the complexities of viruses. Any description of virus inactivation kinetics by this approach is not possible.

IFCT-Mediated TiO 2 /Cu Bacterial/Virus Inactivation
TiO 2 /Cu inactivation of pathogens was first reported by a Sunada et al. [32]. Later, several laboratories reported Cu ions/Cu composites leading to E. coli inactivation by Fenton-like reactions and ROS radicals and peroxides being generated and partly adsorbed on TiO 2 . Inactivation by Cu x O and TiO 2 /Cu films has been the focus of some recent reports [33][34][35]. TiO 2 /Cu lead to a higher degree of cell inactivation compared to TiO 2 due to the Cu toxicity and the additional ROS generation detected in the solution compared to bare TiO 2 .
Why does doping enhance TiO 2 /Cu photocatalytic disinfection? Cu improves the TiO 2 photocatalytic pathogen inactivation compared with TiO 2 alone considering the following parameters: (a) the biocidal nature of Cu and (b) the shift of the NPs' absorption to the visible region up to 600 nm. TiO 2 , absorbing around 4-4.5% of the solar irradiation, has been widely reported to activate pathogen inactivation processes. The bacterial protein for many pathogens absorbs light at up to 320-325 nm [36,37]; (c) the TiO 2 band gap is made narrower by the added Cu/CuO and (d) the Cu clusters on TiO 2 as electron traps hinder, to some extent, e-/h+ recombination [38], and this occurs in conjunction with the charge transfer between the TiO 2 /Cu surfaces and the bacteria. A scheme for the IFCT processes taking place is shown below in Figure 5.
A simplified mechanism for bacterial/virus inactivation is suggested for CuO NPs in Equations (16)-(23) [33,34]. By the use of photon energies exceeding the CuO band gap, the photogenerated cbe-can react with O 2 , leading to the ROS O 2 •− or alternatively reducing the Cu 2+ to Cu + as noted below: CuO(Cu + ) → CuO vacancy + Cu + Sequentially, binary sputtered films of Cu on polyethylene terephthalate CuOx/TiO 2 -PET under visible light promote the interfacial charge transfer (IFCT) of the cbe-from Cu 2 Ocb to TiO 2 cb, as noted in Equation (21). Then, the TiO 2 vbh+ holes with an appropriate positive potential oxidize bacteria, as shown below by Equation (23).
The bacteriophage f2 has been photocatalytically inactivated under visible light irradiation. The removal efficiency of the virus increased with catalyst concentration, light intensity and temperature and decreased at a higher initial virus concentration. Zheng et al. reported on TiO 2 doped with non-noble metals able to inactivate virus under light irradiation [39]. Until now the role of metal-ions in photocatalytic viral disinfection have not been explored in a systematic, comprehensive and detailed way. Composite material semiconductors containing Ag and to a lesser extent Cu have shown to form plasmon under UV-visible light irradiation [40][41][42]. The plasmons play a significant role in bacteria and virus inactivation. This remains a task of the future and seems necessary to increase the virus inactivation kinetics. This is one of the critical issues hindering the commercial scale application of photocatalytic technology for disinfection purposes [43].  Table 1 below. The parentheses show the number of experimental runs performed to report the median value.  It is difficult to differentiate the toxicity introduced individually by Ag-Cu NPs' surfaces and the separate effects induced by Ag and Cu NPs by themselves under light in aqueous media [8,16,25]. Recent studies indicated that Cu NPs induced oxidative damage to bacteria with higher kinetics compared to Ag NPs [24]. Figure 6 suggests the IFCT for Ag2O/CuO composite films under light irradiation and the subsequent generation of radicals, leading to the degradation of organics (RH) or bacteria.

IFCT-Mediated Ag-Cu Bacterial/Virus Inactivation
Studies on Ag-Cu NP films show higher bacterial inactivation kinetics compared to either Ag or Cu alone. Studies on Ag-Cu NP films leading to the destruction of bacteria have been reported recently [44,45]. The composition and species found in these composites are shown in Figure 4.
The faster kinetic inactivation of pathogens by bimetallic/trimetallic oxides is used due to two main reasons: (a) it increases the absorption of semiconductors such as TiO 2 into the visible range and (b) it also increases the number of potential couples available to catalyze/photocatalyze chemical transformations on the catalyst surface. Pathogen inactivation kinetics by bimetals such as Ag-Cu films compared to either Ag or Cu films by themselves. Work on Fe-doped/decorated wide-band semiconductors such as TiO 2 , mediating bacterial inactivation, has been reported recently [1]. Using fast kinetics spectroscopy, the short-lived precursors leading to bacterial inactivation have been sorted out [43,46]. to the visible region up to 600 nm. TiO2, absorbing around 4-4.5% of the solar irradiation, has been widely reported to activate pathogen inactivation processes. The bacterial protein for many pathogens absorbs light at up to 320-325 nm [36,37]; (c) the TiO2 band gap is made narrower by the added Cu/CuO and (d) the Cu clusters on TiO2 as electron traps hinder, to some extent, e-/h+ recombination [38], and this occurs in conjunction with the charge transfer between the TiO2/Cu surfaces and the bacteria. A scheme for the IFCT processes taking place is shown below in Figure 4. A simplified mechanism for bacterial/virus inactivation is suggested for CuO NPs in Equations (16)-(23) [33,34]. By the use of photon energies exceeding the CuO band gap, Killing bacteria on metallic copper surfaces occurs primarily due to the release of copper ions. However, it also involves slower processes such as the contact of the bacteria with the Cu/CuO surface [35,36]. Copper cations participate in Fenton-like reactions, leading to fast cellular damage by ROS species in mammalian cell membranes. This is due to the high Cu cation cytotoxicity. There are no standard tests to evaluate NPs' cytotoxicity which may allow to compare the results found in different laboratories [47][48][49]. The translocation of Cu NPs/Cu ions and Ag ions into the pathogen cytoplasm is far from being understood at the present time. The relative index of the toxicity on mammalian cells of Ag NPs and CuO NPs is shown in Table 1 below. The parentheses show the number of experimental runs performed to report the median value. It is difficult to differentiate the toxicity introduced individually by Ag-Cu NPs' surfaces and the separate effects induced by Ag and Cu NPs by themselves under light in aqueous media [8,16,25]. Recent studies indicated that Cu NPs induced oxidative damage to bacteria with higher kinetics compared to Ag NPs [24]. Figure 6 suggests the IFCT for Ag 2 O/CuO composite films under light irradiation and the subsequent generation of radicals, leading to the degradation of organics (RH) or bacteria.
Like many other metal film composites, Ag-Cu films accelerate the inactivation of pathogens [50,51]. This presents an alternative to the use of antibiotics without significantly increasing the risk of resistance when antibiotics are administered for long periods of time. The efficiency of these films is dependent on the coating technique used to prepare them. A more advanced nanotechnological approach to prepare uniform and reproducible NP polymer films would benefit the inactivation of pathogens within acceptable times [52,53]. The quantitative contribution of each NP to the production of reactive oxygen species is still an open issue. Cu(II) ions are reduced to Cu(I) ions by the enzymes regulating the human respiratory cycle [35]. An excess of copper cations cannot be removed by these enzymes [36] and leads to respiratory cycle disfunction and death within short times.
Cu ions 53 (10) It is difficult to differentiate the toxicity introduced individually by Ag-Cu NPs' surfaces and the separate effects induced by Ag and Cu NPs by themselves under light in aqueous media [8,16,25]. Recent studies indicated that Cu NPs induced oxidative damage to bacteria with higher kinetics compared to Ag NPs [24]. Figure 6 suggests the IFCT for Ag2O/CuO composite films under light irradiation and the subsequent generation of radicals, leading to the degradation of organics (RH) or bacteria.  Surface plasmon resonance (SPR) bands have been investigated for some years for Au, Ag and Cu nanoparticles. Cu SPR films are also applied in optical catalysis, detection and conductive devices used in printing technologies [52]. Until now, no comprehensive mechanism has been reported for the detailed intervention of Cu plasmons leading to bacterial/virus inactivation, and Au plasmons intervening in bacterial inactivation have recently been reported [43]. Visible light irradiation of SPR Au/TiO 2 bands activates the electron injection of Au NPs into TiO 2 with a slight dissolution of the Au. This dissolution leads to (a) the injection of hot Au electrons into TiO 2 followed by femtosecond spectroscopy and (b) the generation of Au ions in the solution [43]. This study also reports on Au ions' mechanism of interaction with bacterial cells. The bactericidal inactivation mechanism by Au ions is significantly different to the mechanism reported for Ag SPR and Cu SPR activated by light.
Higher toxicity has been extensively reported for Cu relative to Ag. Extremely low amounts of Cu species (in the nano range) induce pathogen inactivation in the dark due to the high cytotoxicity of Cu per unit weight compared to Ag. Cu is not a noble metal and is readily available compared to Ag. This makes Cu/CuO/Cu ions a preferred choice for pathogen inactivation studies. Copper is number 29 of the periodic table of elements. The Cu atomic structure is made up of negatively charged electrons in completely filled orbitals close to the positively charged atomic nucleus. The electrons of the unfilled orbitals 4s 1 3d 10 require a small input of energy to activate chemical reactions. Cu(I) (4s 0 3d 10 ) and Cu(II) (4s 0 3d 9 ) oxidation states require a much lower one-electron redox potential compared to Ag(I)/Ag(II) and Fe(II)/Fe(III) in electron transfer reactions. Gram-negative bacteria including Cu in the periplasm can handle the toxicity of copper cations up to a certain level in their own metabolism.  Figure 7 presents a schematic for the Fe 2 O 3 -TiO 2 intervention [54]. The conduction band (cb) of Fe 2 O 3 is located at a lower potential energy level with respect to TiO 2 , as shown in Figure 7. However, electron transfer is possible since the TiO 2 lattice presents trapping sites~0.8 eV below the position of the TiO 2 cb [4,55,56]. The charge transfer between the two semiconductors led to (a) an increase in the photogenerated charge separation in Fe 2 O, (b) an increase in the pollutant oxidation kinetics by the generated holes and (c) improved reduction kinetics of the (O 2 ) air by the photogenerated (cb) electrons. The TiO 2 vb generates the • OH radicals leading to bacterial/virus inactivation due to the more favorable potential energy position of its vb. The charge transport in the Fe 2 O 3 -TiO 2 controls the photocatalytic-mediated reaction kinetics [57][58][59][60]. Under light, the Fe 2+ in the Fe 2 O 3 -TiO 2 undergoes a redox reaction and oxidizes to Fe 3+ , as noted in Equation (24)   The schema in Figure 8 suggests that with the FeOx/TiO2-PE, bacterial inactivation on the film is possible from the Fe oxide into TiO2 due to low-lying states detected below the conduction band (cb) of TiO2. This allows to suggest a scheme of reactions leading to a short-lived unstable bacteria cation(+) as noted below in Equation (25) and reported in reference [43]. During the bacterial abatement process shown in Figure 7, some Fe ions leach out of the FeOx-TiO2-PE films under light irradiation in an aqueous solution. This gives rise to Fenton-like reactions with ions such as Fe, Mn or Cu ions [20]. ROS radicals are generated in solution [20]. The reaction sequence leading to virus/bacterial inactivation proceeds The schema in Figure 8 suggests that with the FeOx/TiO 2 -PE, bacterial inactivation on the film is possible from the Fe oxide into TiO 2 due to low-lying states detected below the conduction band (cb) of TiO 2 . This allows to suggest a scheme of reactions leading to a short-lived unstable bacteria cation(+) as noted below in Equation (25) and reported in reference [43].

IFCT in Fe 2 O 3 /TiO 2 Giving Rise to Fenton-Like Reactions Mediating Virus/Bacterial Inactivation
During the bacterial abatement process shown in Figure 7, some Fe ions leach out of the FeOx-TiO 2 -PE films under light irradiation in an aqueous solution. This gives rise to Fenton-like reactions with ions such as Fe, Mn or Cu ions [20]. ROS radicals are generated in solution [20]. The reaction sequence leading to virus/bacterial inactivation proceeds through several steps as detailed next: (a) virus adsorption on the catalyst; (b) ROS generation and surface and bulk reactions and (c) desorption of the inactivated viruses. The diffusion of the virus within the inoculate plays a role in the inactivation process [9]. Mononuclear metalloenzymes include Fe(II) in their structure since Fe ions are important in human/animal metabolism. These mononuclear Fe enzymes are made up of Fe and Zn(II) and Mn(II). Fenton chemistry occurs when FeOx corrodes, leading to Fe ions in solution [61]. Fe enzyme Fe-S dehydratase such as Fe(II)-cysteine is oxidized through Fenton reactions to the sulphinic RSO(=H) or sulphonic RS=O 2 -OH species, displacing Fe(II) from the metalloenzyme. The latter reaction involves several steps: (a) oxidant depletion and (b) additional (excess) ROS and thiol depletion from the cell bilayer topmost functional groups [62]. Thiol depletion due to Fenton-like reactions is shown schematically in Figure 9. The outcome is that normal cell metabolism is inhibited, leading to cell death. Fe-S-type dehydratases are particularly sensitive to the destruction of the topmost bilayer functional binding groups containing S-moieties [63]. The schema in Figure 8 suggests that with the FeOx/TiO2-PE, bacterial inactivati on the film is possible from the Fe oxide into TiO2 due to low-lying states detected belo the conduction band (cb) of TiO2. This allows to suggest a scheme of reactions leading a short-lived unstable bacteria cation(+) as noted below in Equation (25) and reported reference [43].

Bacteria + [PE-FeOx] + Vis light  [Bacteria*…FeOx] PE  [Bacteria + …FeOx cbe]PE
(2 During the bacterial abatement process shown in Figure 7, some Fe ions leach out the FeOx-TiO2-PE films under light irradiation in an aqueous solution. This gives rise Fenton-like reactions with ions such as Fe, Mn or Cu ions [20]. ROS radicals are generat in solution [20]. The reaction sequence leading to virus/bacterial inactivation procee  [62]. Thiol depletion due to Fenton-like reactions is shown schematically in Figure 9. The ou come is that normal cell metabolism is inhibited, leading to cell death. Fe-S-type dehydratase are particularly sensitive to the destruction of the topmost bilayer functional binding group containing S-moieties [63]. , a quasi-stationary state, since the concentration of OH . is very small. The OH radical is generated in sub-nanomolar concentrations as shown by the equation ( • OH) = k 1 (Fe 2+ )(H 2 O 2 )/k 2 (RH) in quantities of about~10 −10 M. This latter value corresponds to a sub-nanomolar concentration. The concentration estimated for the OH radicals is similar to the concentration of OH radicals found in many natural water bodies.

IFCT in Ternary Semiconductors Leading to Bacterial/Virus Inactivation
A mechanism is suggested in Figure 10a for the interfacial charge transfer (IFCT) in the Ag(3%)-TiO 2 -FeO x (3%) photocatalyst mediating bacterial inactivation [64]. The TiO 2 cb position in Ag(3%)-TiO 2 -FeO x (3%) is more cathodic compared to FeO x cb, and the TiO 2 vb is more anodic in respect to FeO x . The potential energy positions of the potential electronic bands of the three oxides determine the IFCT kinetics taking place in Figure 10a. The release of Fe-ions in aqueous solutions by the magnetic Ag(3%)-TiO 2 -FeO x (3%) composite has been reported for several years when using Fe-composites. In the case of the Ag(3%)-TiO 2 -FeO x (3%), a high calcination temperature was used during the preparation of this composite. The calcination step strongly binds the Fe to the TiO 2 through Ti-OH bonds. This precludes the Fe-ions release into solution within the bacterial inactivation process. The FeO x (mainly Fe 2 O 3 ) in the Ag(3%)-TiO 2 -FeO x (3%) during the calcination step becomes deeply buried in the catalyst bulk. If present in the surface, the FeOx would photo-corrode in aqueous media under light irradiation [65]. The significant acceleration in the bacterial inactivation kinetics by Ag(3%)-TiO 2 -FeO x (3%) is due to the localized electric field leading to more reactions per unit time compared to the non-magnetized catalyst Ag(3%)-TiO 2 . This leads to the polarization in the Ag(3%)-TiO 2 -FeO x (3%) heterojunction enhancing the separation of electrons and holes [66]. Repetitive recycling of the Ag(3%)-TiO 2 -FeO x (3%) did not lead to loss of catalytic activity. This provides evidence for the stable nature of the FeO x present in the composite photocatalyst.
The use of heterostructures made up by wide-band semiconductors, oxide and metals reduces the band gap of TiO 2 (bg 3.2 eV). As a result, development of visible light active titanium dioxide materials was observed. This is one of the main challenges in the field of semiconductor photocatalysis. Heterostructures composed of ternary catalysts, doped or not, enhance light absorption in the visible region. Figure 10b shows an example of IFCT over ternary semiconductor photocatalysts. More work to identify appropriate dopants to improve visible light absorption and electron-hole separation is needed at this time to enhance photocatalytic activity. One serious problem encountered in binary or ternary semiconductors designed to accelerate bacterial/virus inactivation is the electrostatic recombination of the cbe-of one semiconductor with the vbh+ holes of the other semiconductor. This hinders the diffusion of photogenerated charges to the semiconductor surface, lowering the production of readily available photogenerated species [67].
Semiconductors have many advantages over nanoparticle dispersions or colloidal dispersions in photocatalytic reactions activated by solar/visible light irradiation. Long et al. [84] reported on the beneficial effect of intrinsic structural defects in a semicon-ductor on their chemical reactivity. This affects the stability and electronic properties of semiconductors. In some cases, these defects enhance the potential energy position of the semiconductor electronic bands. Fu et al. recently reported that the photocatalytic efficiency of a semiconductor varied with the type of defects in the composite, whether it was a surface, bulk or adsorbate defect. Defects modify the semiconductor band bending and, therefore, the semiconductor electronic structure. Both factors introduce changes in the semiconductor IFCT process mediating the chemical transformation. The potential to improve semiconductor performance by the introduction of adequate defects in the semiconductor structure is important for any future application of semiconductor devices.

Viral Biofilms
Biofilms are aggregates of bacteria/viruses on a surface surrounded by a protective coating that keep for long times on surfaces. They have been recognized as the most effective way of spreading dangerous pathogens into the environment. Biofilms induce diseases that account for more than 80% of viral/bacterial infections. The biofilms once anchored on supports are resistant to their removal by conventional antiseptics or disinfect-

Viral Biofilms
Biofilms are aggregates of bacteria/viruses on a surface surrounded by a protective coating that keep for long times on surfaces. They have been recognized as the most effec-tive way of spreading dangerous pathogens into the environment. Biofilms induce diseases that account for more than 80% of viral/bacterial infections. The biofilms once anchored on supports are resistant to their removal by conventional antiseptics or disinfectants. Viruses are capable of forming complex biofilm-like assemblies, similar to bacterial biofilms, and disseminate viruses dangerous pathogens into the environment for long periods of time, especially within hospitals and healthcare facilities, depending on local conditions. Biofilms are usually resilient to stress, lack of water, pH effects and mechanical stress or impacts. The stages of biofilm formation are schematically suggested in Figure 11 [85].
(a) Biofilm formation: The first stage in biofilm formation is the adhesion of a virus on a surface by adsorption, involving hydrophobic effects through covalent bonding. Roughness of the surface favors this step and avoids liquids flowing near the biofilms to preclude biofilm formation. The second stage involves bacteria reproduction to form a colony matrix, with this step being concomitant to the growth of an extracellular polymeric shell protecting the colony. In the third/last step, the colony attains its critical mass, ingests nutrients and eliminates metabolic residuals with a kinetics regulated by its enzymes [86]. (b) The degradation of recalcitrant biofilms occurs in different ways and is suggested next in Figure 12, involving several steps: (a) surface structural modification to preclude biofilm adhesion; (b) the use of bactericidal agents inducing quorum quenching/enzymatic/immunological disruption and (c) the use of catalysts under light or in the dark, leading to bacterial interference in the biofilm [87]. Biofilm destruction is important since many films are resilient to degradation by antibiotic metal/oxide or chloro-compounds. Human coronavirus (HCoV) films lead to respiratory diseases [38]. to preclude biofilm formation. The second stage involves bacteria reproduction to form a colony matrix, with this step being concomitant to the growth of an extracellular polymeric shell protecting the colony. In the third/last step, the colony attains its critical mass, ingests nutrients and eliminates metabolic residuals with a kinetics regulated by its enzymes [86]. (b) The degradation of recalcitrant biofilms occurs in different ways and is suggested next in Figure 12, involving several steps: (a) surface structural modification to preclude biofilm adhesion; (b) the use of bactericidal agents inducing quorum quenching/enzymatic/immunological disruption and (c) the use of catalysts under light or in the dark, leading to bacterial interference in the biofilm [87]. Biofilm destruction is important since many films are resilient to degradation by antibiotic metal/oxide or chloro-compounds. Human coronavirus (HCoV) films lead to respiratory diseases [38]. TiO2/Ag, TiO2/Cu and TiO2/Fe2O3 composites under light irradiation involving IFCT processes have been discussed above in this study and illustrated with a few relevant examples. Figure 11. Growth of a virus biofilm. Initially, viruses settle down onto a surface that can either be inanimate or living tissue. Then, secretion of extracellular matrix allows the cells to colonize and anchor on the substrate. This step repeats itself, leading to biofilm growth. Finally, a fully developed biofilm with water channels allows oxygen and nutrients to penetrate to the biofilm and allows processes disseminating viruses into the environment for a long time. Figure 11. Growth of a virus biofilm. Initially, viruses settle down onto a surface that can either be inanimate or living tissue. Then, secretion of extracellular matrix allows the cells to colonize and anchor on the substrate. This step repeats itself, leading to biofilm growth. Finally, a fully developed biofilm with water channels allows oxygen and nutrients to penetrate to the biofilm and allows processes disseminating viruses into the environment for a long time. Figure 11. Growth of a virus biofilm. Initially, viruses settle down onto a surface that can either be inanimate or living tissue. Then, secretion of extracellular matrix allows the cells to colonize and anchor on the substrate. This step repeats itself, leading to biofilm growth. Finally, a fully developed biofilm with water channels allows oxygen and nutrients to penetrate to the biofilm and allows processes disseminating viruses into the environment for a long time.

Conclusions and Outlook for Future Work
This update briefly describes the scientific bases for the use of semiconductor and metal/semiconductor materials, addressing critical issues in the photocatalytic pathogen inactivation field. Some fundamental issues related to the inactivation of pathogens by materials under visible light irradiation were summarily described. The advantages and limitations of these innovative materials leading to pathogen inactivation are briefly described. The advantages of photochemical processes in bacterial/virus disinfection compared to the use of chlorinated compounds were addressed in detail. The objective of this update was to guide the readership working in microbiology and photocatalysis on disinfection by a mild and environmentally friendly technique.
Future work should address the synthesis of innovative, better performing catalysts. These materials may be applied in conjunction with oxidants commonly present in water reservoirs such as H 2 O 2 /organic peroxides. The virus inactivation kinetics of supported films has been reported to be below the one found for clusters/colloids/powders. This is due to the lower surface area available in the film compared to dispersions. However, microparticles dissolved in solution have to be separated after the disinfection process. This is not the case for bactericide films. Films presenting high adsorption for bacteria/viruses could remove this disadvantage and enable their use in photoreactors, useful in continuous processes.
Little is known, to date, on the subject of the viral disinfection kinetics mechanism. Virus disinfection kinetics does not follow the first-order kinetics reported for organic pollutant degradation induced by TiO 2 under light irradiation. This is due to the more complicated viral structure, envelope and repair mechanisms against ROS damage. Much of the present work to inactivate viruses by photocatalytic treatment follows the approach used to inactivate bacteria. However, the structure, surface, size, hydrophobic balance and redox properties of viruses are quite different to those found in bacteria. Innovative materials should be biocompatible, stable and present during long operational times and should present an acceptable corrosion resistance. This enables the release of NPs into solution within the threshold fixed by sanitary regulations. Novel Cu disinfection materials precluding Cu release from the matrix should be designed but should still present contact catalytic sites on their surface that are able to inactivate pathogens. A precedent for this type of material is the case of Fe catalysts/photocatalysts, proceeding without the release of Fe-from TiO 2 matrixes [54]. Future research should address the synthesis of more advanced double metal oxides and other composites made up of biocompatible metal/ions with high redox potentials such as Ag, Mn, Fe and Au compared to traditional semiconductors such as TiO 2 , ZnO and MnO 2 . Controlled release of metals/ions in sub-nanomolar amounts into water bodies has not been addressed, until now, to induce virus inactivation.