Digitally Printed AgNPs Doped TiO2 on Commercial Porcelain-Grès Tiles: Synergistic Effects and Continuous Photocatalytic Antibacterial Activity

In the present study, we use commercial digitally printed ceramic tiles, functionnalized by AgNPs doped micro–TiO2, to investigate the mechanism of Ag in the continouos photocatalytic antibacterial activity. The novelty of the research lies in the attempt to understand the mechanism of Ag, supported on TiO2, able to exhibit the same antibacterial activity of a standard system containing Ag species, but here, totally embedded on the tile surface, and thus not free to move and damage the bacteria cell. UV/vis diffuse reflectance spectroscopy (DRS) of AgNPs–TiO2 tiles indicated an enhanced visible light response, wherein a new absorption band was produced around 18,000–20,000 cm−1 (i.e., in the 400–600 nm range) owing to the surface plasmon resonance (SPR) of AgNPs. The antibacterial photocatalytic experiments were conducted towards the inactivation of E. coli under solar light and indoor light. It was found that the degradation speed of E. coli in the presence of AgNPs–TiO2 tiles is solar light-intensity depending. This justifies the semiconductor behavior of the material. Furthermore, the AgNPs–TiO2 tiles exhibit a high ability for the inactivation of E. coli at a high load (104–107 colony-forming unit (CFU)/mL). Additionally, AgNPs–TiO2 tiles showed a remarkable antibacterial activity under indoor light, which confirms the good photocatalytic ability of such tiles. On the basis of the reactive oxygen species (ROS) quenching experiments, O2•− species and h+ were more reactive for the inactivation of E. coli rather than •OH species. This is because of the different lifetime (bacteria are more likely oxidized by ROS with longer lifetime); in fact, O2•− and h+ exhibit a longer lifetime compared with •OH species. The generation of H2O2 as the most stable ROS molecule was also suggested.

ions and AgNPs to the cell wall and core. Notably, in this study, we investigate a new antibacterial mechanism of the as-prepared material. In fact, the core novelty of our study is the development of an antibacterial material characterized by AgNPs blocked on the semiconductor surface. XPS analysis showed the co-existence of both Ag + and metallic Ag. In particular, the as prepared porcelain Ag-TiO 2 printed surfaces were tested under solar and visible light irradiation for antimicrobial applications on Gram-negative pathogenic bacteria such as E. coli. Finally, stereomicroscopy analysis showed dead bacteria within 180 min. With appropriate and specially chosen quenchers, the bacterial inactivation mechanisms under solar and visible light are proposed.

Materials
The commercial 1077-Kronos was used as a photoactive semiconductor to design the photoactive tiles. It is pure anatase phase and the main features of this well-known material are summarized in Table 1. Silver nitrate (AgNO 3 , ACS Reagent, ≥99%) was used as metal precursor. The other reagents were polyvinylpyrrolidone (PVP40, average mol wt −1 : 40,000) and sodium borohydride (NaBH 4 , ≥99%), all Sigma Aldrich reagents used without further purification or pre-treatments.

Synthesis of Anti-Bacterial Ag@TiO 2 Photoactive Porcelain-Grès Tiles
The detailed description of the synthesis of Ag@TiO 2 photoactive tiles was reported in our previous studies [18,19]. Three different samples with 1%, 4%, and 8% wt AgNPs on micro-sized TiO 2 were prepared. Typically, the surface of commercial photoactive porcelain-grès tiles is covered with a formulation containing pure anatase micro-TiO 2 (1077 by Kronos) doped with 8% AgNPs from a silver nitrate solution. The digital printing production is applied for the preparation of the functionalized industrial porcelain-grès tiles (Figure 1a). A digital printer (Projecta S.p.A., Italy) impressed the solvent-based ink containing the photocatalyst on tiles surface. After the ink deposition, tiles were calcined at 680 • C for 80 min in a kiln, with high-controlled temperature ramp avoiding thermal shock, and washed with water with a brushing system to remove the particle of non-adherent TiO 2 . The water used for degreasing was recovered and purified prior to recycling and any damaged tiles were recycled [5]. The main advantages of this procedure are reported as follows: green production with no waste of photocatalytic powder (TiO 2 ), no use of water, smaller consumption of energy, CO 2 emissions lower than 89%, possibility to work on large slabs (300 cm × 150 cm).
A remarkable great stability of the photocatalytic coating was achieved. To quickly test the stability of the washcoat adherence, the coated tile was immersed in a vessel containing liquid nitrogen; no detachment of photocatalytic material was observed despite the drastic conditions ( Figure 1b).

Characteristics of Ag@TiO2 Photoactive Porcelain-Grès Tiles
The material was characterized by means of many different experimental techniques, including HR-SEM, HR-TEM, EDX, and XRD. The detailed description of the performed analyses is reported elsewhere [18]. However, from the EDX analysis, it was possible to determine the amount of TiO2 coated on the tile surface as 1.1 g m −2 . SEM showed a good distribution of the micro-TiO2 on the porcelain tile surface prepared by the digital printing process.
More details and discussion regarding the characterization of AgNPs-TiO2 photoactive porcelain-grès tiles are reported in our previous research [18,19]. UV/visible diffuse reflectance spectra (UV-DRS) absorption responses of AgNPs-TiO2 samples with different amount of AgNPs loading were obtained by means of a UV/vis scanning spectrophotometer (PerkinElmer, Lambda 35 model, Waltham, MA, USA), equipped with a diffuse reflectance accessory. The possible co-existence of Ag + and metallic Ag was detected by XPS analysis (M-Probe, SSI Instrument, USA).

Photocatalytic Properties
Two different light sources were used during the course of this study, indoor actinic light and solar simulated light. The indoor light irradiation of the samples was carried out in a cavity provided with tubular Osram Lumilux 18W/827, Munich, Germany. These lamps emitted in the range of 340-720 nm with an integral output of 5.0 mW/cm 2 . A solar simulator for Hereaus GmbH (Hanau, Germany) was used to irradiate the samples with three different irradiances (100, 50, and 20 mW/cm 2 ). The solar simulator emits from a 1500 W Xenon lamp at wavelengths between 310 and 750 nm; a quartz glass cut-off filter cuts the other wavelengths (UVB, UVC, and IR).
The light intensity was monitored using a UV radiometer and a global irradiance couple (CUV3 and CM3 Models, Kipp & Zonen B.V., Delft, The Netherlands).

Bacterial Inactivation and Used Light Sources
The samples of Escherichia coli (E. coli) were obtained from the Deutsche Sammlung von Mikroorganismen und Zellkulturen GmbH (DSMZ) ATCC23716, Braunschweig, Germany, to test the antibacterial activity of the Ag-coated tiles. Prior to testing, all tiles were autoclaved at 121 °C for 2 h. The bacterial aliquots in NaCl/KCl (pH 7.2) were placed on coated and uncoated (control) tiles. Because the tiles present a micro-porosity, an adsorption stage of 30 min was necessary in order to guarantee an optimal contact between the bacterial cells and the reactive interface. A 100 μL E. coli inoculum was contacted with each sample. The inoculation was done at room temperature (25-28 °C). This temperature did not vary much during the bacterial testing. The tiles/samples were then placed in Petri dishes provided with a lid to prevent evaporation, and then exposed to light for the

Characteristics of Ag@TiO 2 Photoactive Porcelain-Grès Tiles
The material was characterized by means of many different experimental techniques, including HR-SEM, HR-TEM, EDX, and XRD. The detailed description of the performed analyses is reported elsewhere [18]. However, from the EDX analysis, it was possible to determine the amount of TiO 2 coated on the tile surface as 1.1 g m −2 . SEM showed a good distribution of the micro-TiO 2 on the porcelain tile surface prepared by the digital printing process.
More details and discussion regarding the characterization of AgNPs-TiO 2 photoactive porcelain-grès tiles are reported in our previous research [18,19]. UV/visible diffuse reflectance spectra (UV-DRS) absorption responses of AgNPs-TiO 2 samples with different amount of AgNPs loading were obtained by means of a UV/vis scanning spectrophotometer (PerkinElmer, Lambda 35 model, Waltham, MA, USA), equipped with a diffuse reflectance accessory. The possible co-existence of Ag + and metallic Ag was detected by XPS analysis (M-Probe, SSI Instrument, USA).

Photocatalytic Properties
Two different light sources were used during the course of this study, indoor actinic light and solar simulated light. The indoor light irradiation of the samples was carried out in a cavity provided with tubular Osram Lumilux 18W/827, Munich, Germany. These lamps emitted in the range of 340-720 nm with an integral output of 5.0 mW/cm 2 . A solar simulator for Hereaus GmbH (Hanau, Germany) was used to irradiate the samples with three different irradiances (100, 50, and 20 mW/cm 2 ). The solar simulator emits from a 1500 W Xenon lamp at wavelengths between 310 and 750 nm; a quartz glass cut-off filter cuts the other wavelengths (UVB, UVC, and IR).
The light intensity was monitored using a UV radiometer and a global irradiance couple (CUV3 and CM3 Models, Kipp & Zonen B.V., Delft, The Netherlands).

Bacterial Inactivation and Used Light Sources
The samples of Escherichia coli (E. coli) were obtained from the Deutsche Sammlung von Mikroorganismen und Zellkulturen GmbH (DSMZ) ATCC23716, Braunschweig, Germany, to test the antibacterial activity of the Ag-coated tiles. Prior to testing, all tiles were autoclaved at 121 • C for 2 h. The bacterial aliquots in NaCl/KCl (pH 7.2) were placed on coated and uncoated (control) tiles. Because the tiles present a micro-porosity, an adsorption stage of 30 min was necessary in order to guarantee an optimal contact between the bacterial cells and the reactive interface. A 100 µL E. coli inoculum was contacted with each sample. The inoculation was done at room temperature (25-28 • C). This temperature did not vary much during the bacterial testing. The tiles/samples were then placed in Petri dishes provided with a lid to prevent evaporation, and then exposed to light for the photocatalytic activity testing. After each determination, the tiles were transferred into a sterile tube containing 5 mL autoclaved saline solution. This solution was subsequently mixed thoroughly using a vortex for 2 min. Serial dilutions were made in NaCl/KCl solution. A 100 µL sample of each dilution was pipetted onto a nutrient agar plate and then spread over the surface of the plate using the standard plate method. These agar plates were then incubated lid down at 37 • C for 24 h before colonies were counted. Three independent assays were done for each sputtered sample. The statistical analysis of the results was performed for the CFU values, calculating the standard deviation values (n = 5%).
To verify that no re-growth of E. coli occurs after the first bacterial inactivation cycle, the tiles were incubated for 24 h at 37 • C. These samples were incubated at 37 • C for 24 h. Bacterial regrowth was observed only when using indoor light.

Stereomicroscopy and Live/Dead Bacteria
The stereomicroscopy imaging was carried out on samples inoculated with 6 × 10 6 CFU/mL of E. coli and incubated for 2 h in a humidification chamber. This method uses a fluorochrome ® Biofilm Viability Kit (Molecular Probes, Invitrogen, ThermoFisher Instruments, Waltham, Massachusetts USA). The kit contains a combination of the SYTO9 ® green fluorescent nucleic acid stain and propidium iodide (PI) fluorochromes for the staining of live and dead cells, respectively. The sample fluorescence was monitored in a stereomicroscope (Leica MZ16 FA, Leica Microsystems GmbH Wetzlar, Germany) and the images were processed using the Leica Application Suite LAS v.1.7.0 build 1240 software from Leica Microsystems CMS GmbH (Wertzlar, Germany). Adhesion of bacteria on the tiles was allowed for 15 min before washing the sample with sterile Milli-Q water to remove non-adherent bacteria.

Micro-Oxidation (Local pH) and Interfacial Potential at the Surface of the Coated Tiles
The local pH and interfacial potentials were monitored during the time of bacterial inactivation by means of a Jenco 6230N (pH/mV/temperature meter) with a hand-held microprocessor in splash proof case with a three-point calibration. The signals were monitored via RS-232-C IBM compatible communication interface and BNC, pH/ORP, and eight-pin DIN ATC connector. The redox/pH electrode was standardized using two redox solutions (240 mV and 470 mV; Hanna Instruments, Switzerland) at three pH standards (3, 8, and 10). After each analysis, the electrodes were sterilized in a mixture (70% v/v) of distilled water and ethyl alcohol (96% v/v) for 16 h. Before each experiment, the electrodes were washed with sterilized MilliQ-water. To maintain the electrode performance, the electrodes were polished after each run with a fine alumina oxide powder (Sigma Aldrich, Buchs Switzerland).

ROS Quenchers
Aliquots of the bacterial suspension were suspended on the ceramic tiles in either the absence (control) or presence of chelating agents (ROS quenchers/scavengers), adding a 50 µL of bacterial suspension at a high concentration (10 6 CFU/mL). Ethylenediamine-tetra-acetic acid sodium (EDTA-2Na) was added as hole (h + ) scavenger. Dimethylsulfoxide (DMSO) was used as ( • OH) scavenger and superoxide dismutase (SOD) as O 2 •− scavenger. A concentration of 0.2 mM of scavenger was added in each case to identify each intermediate/radical species.

Photocatalyst Characterization
The morphology of Ag@TiO 2 deposited into porcelain-grès tiles was characterized by SEM analysis, as shown in Figure 2a, wherein small sized nanoparticles of Ag@TiO 2 were aggregated on the surface, showing a uniform coating. For further clarification, HR-TEM analysis was recorded on bare-micro TiO 2 and Ag@TiO 2 (Figure 2b,c). Bare TiO 2 exhibits a highly ordered structure and the inspection of the fringe patterns confirms the presence of the anatase phase (mostly exposed family planes (101) of anatase [ICDD card n. ). In Ag@TiO 2 samples, AgNPs are clearly evident on the surface of TiO 2 , exhibiting different particles' dimensions in the 3-20 nm range. Both the presence of anatase as unique TiO 2 polymorph, for either the bare support and/or all Ag-promoted materials, and the presence of Ag itself have been confirmed by XRD measurements; no figure has been reported for the sake of brevity, as data have been already published elsewhere [20].
Surfaces 2020, 3 FOR PEER REVIEW 6 inspection of the fringe patterns confirms the presence of the anatase phase (mostly exposed family planes (101) of anatase [ICDD card n. ). In Ag@TiO2 samples, AgNPs are clearly evident on the surface of TiO2, exhibiting different particles' dimensions in the 3-20 nm range. Both the presence of anatase as unique TiO2 polymorph, for either the bare support and/or all Ag-promoted materials, and the presence of Ag itself have been confirmed by XRD measurements; no figure has been reported for the sake of brevity, as data have been already published elsewhere [20]. In band gap engineering, the doping of TiO2 by cation or anions species is a great approach for enhanced visible light response (bathochromic shift), via two possible mechanistic pathways [21] including either the minimizing of the band gap or the formation of intra-band gap states above the valence band. AgNPs can act as electron receivers and catalytic redox sites in an Ag-doped micro-TiO2 system. Photo-excited electrons of micro-TiO2 are likely transferred from the conduction band of TiO2 to AgNPs because the Fermi levels of Ag are lower than that of TiO2, while the lifetime of oxidative positive holes became longer [22]. This mechanism leads to inhibition of the recombination of electron/hole charges, which in turn enhance the photocatalytic activity. Above all, however, the photocatalytic activity of TiO2 under visible light is significantly boosted when it is assisted by surface plasmon resonance (SPR) [23], wherein the photoexcited electrons under visible light of AgNPs are injected to the conduction band (CB) of TiO2.
UV/visible absorption spectra of AgNPs-TiO2 samples ( Figure 3) indicate that the increase of AgNPs amount in TiO2 has almost no effect on the absorption edge of TiO2. On the contrary, the band relative to the Ag plasmon moves towards lower wavenumbers (from 20,600 cm −1 for Ag1% to 18,500 cm −1 for Ag8%) as the Ag content increases; moreover, the intensity of the plasmon band In band gap engineering, the doping of TiO 2 by cation or anions species is a great approach for enhanced visible light response (bathochromic shift), via two possible mechanistic pathways [21] including either the minimizing of the band gap or the formation of intra-band gap states above the valence band. AgNPs can act as electron receivers and catalytic redox sites in an Ag-doped micro-TiO 2 system. Photo-excited electrons of micro-TiO 2 are likely transferred from the conduction band of TiO 2 to AgNPs because the Fermi levels of Ag are lower than that of TiO 2 , while the lifetime of oxidative positive holes became longer [22]. This mechanism leads to inhibition of the recombination of electron/hole charges, which in turn enhance the photocatalytic activity. Above all, however, the photocatalytic activity of TiO 2 under visible light is significantly boosted when it is assisted by surface plasmon resonance (SPR) [23], wherein the photoexcited electrons under visible light of AgNPs are injected to the conduction band (CB) of TiO 2 .
UV/visible absorption spectra of AgNPs-TiO 2 samples ( Figure 3) indicate that the increase of AgNPs amount in TiO 2 has almost no effect on the absorption edge of TiO 2 . On the contrary, the band relative to the Ag plasmon moves towards lower wavenumbers (from 20,600 cm −1 for Ag1% to 18,500 cm −1 for Ag8%) as the Ag content increases; moreover, the intensity of the plasmon band grows up in intensity as the Ag content increases as well as the average size of the Ag particles. Localized SPR of AgNPs can significantly enhance the overall conductivity of TiO 2 and decrease the recombination of electron/hole charges. Furthermore, a sensitizing Ag 2 O with TiO 2 under visible light may take place. The introduction of AgNPs to TiO 2 results in the formation of the Schottky barrier between AgNPs and TiO 2 [24]. The SPR photogenerated electrons under visible light can be injected to the conduction band via this barrier, leading to activating TiO 2 under visible light [25].
Surfaces 2020, 3 FOR PEER REVIEW 7 grows up in intensity as the Ag content increases as well as the average size of the Ag particles. Localized SPR of AgNPs can significantly enhance the overall conductivity of TiO2 and decrease the recombination of electron/hole charges. Furthermore, a sensitizing Ag2O with TiO2 under visible light may take place. The introduction of AgNPs to TiO2 results in the formation of the Schottky barrier between AgNPs and TiO2 [24]. The SPR photogenerated electrons under visible light can be injected to the conduction band via this barrier, leading to activating TiO2 under visible light [25].  Figure 4 shows the high-resolution XPS spectrum of the Ag 4f7/2 and 4f5/2 region. The co-existence of Ag + (Ag2O) and metallic Ag(0) was detected in AgNPs-TiO2 photoactive porcelain-grès tiles [26]. However, it can be noticed that Ag2O is the dominant species compared with metallic Ag.
The main peak at 377 eV is the result of the presence of potassium (K 2s peak), element present as contaminant in the composition of the Kronos K1077 (www.XPSdata.com).   Figure 4 shows the high-resolution XPS spectrum of the Ag 4f 7/2 and 4f 5/2 region. The co-existence of Ag + (Ag 2 O) and metallic Ag(0) was detected in AgNPs-TiO 2 photoactive porcelain-grès tiles [26]. However, it can be noticed that Ag 2 O is the dominant species compared with metallic Ag.
The main peak at 377 eV is the result of the presence of potassium (K 2s peak), element present as contaminant in the composition of the Kronos K1077 (www.XPSdata.com).
Surfaces 2020, 3 FOR PEER REVIEW 7 grows up in intensity as the Ag content increases as well as the average size of the Ag particles. Localized SPR of AgNPs can significantly enhance the overall conductivity of TiO2 and decrease the recombination of electron/hole charges. Furthermore, a sensitizing Ag2O with TiO2 under visible light may take place. The introduction of AgNPs to TiO2 results in the formation of the Schottky barrier between AgNPs and TiO2 [24]. The SPR photogenerated electrons under visible light can be injected to the conduction band via this barrier, leading to activating TiO2 under visible light [25].  Figure 4 shows the high-resolution XPS spectrum of the Ag 4f7/2 and 4f5/2 region. The co-existence of Ag + (Ag2O) and metallic Ag(0) was detected in AgNPs-TiO2 photoactive porcelain-grès tiles [26]. However, it can be noticed that Ag2O is the dominant species compared with metallic Ag.
The main peak at 377 eV is the result of the presence of potassium (K 2s peak), element present as contaminant in the composition of the Kronos K1077 (www.XPSdata.com).

E. coli Inactivation under Different Solar Light Intensities
In order to demonstrate the photocatalytic inactivation of E. coli on the surface of Ag-NPs-TiO 2 tiles, (i) dark run, (ii) photolysis, and (iii) photocatalytic experiments in the presence of Ag-NPs tiles under different light intensities were performed, and the results are shown in Figure 5. It is readily seen that no significant adsorption of E. coli on AgNPs-TiO 2 tiles was found. Furthermore, the photolysis of E. coli without AgNPs-TiO 2 tiles was not obvious within 300 min. However, in the presence of AgNPs-TiO 2 tiles under light, a significant inactivation of E. coli that increases as a function of light intensity increases. This justifies the semiconductor behaviour of the Ag-species at the interface with bacteria [16,27,28]. Thus, these results indicate that the inactivation of E. coli is a light-dependent reaction owing to the formation of ROS on the surface of AgNPs-TiO 2 .

E. coli Inactivation under Different Solar Light Intensities
In order to demonstrate the photocatalytic inactivation of E. coli on the surface of Ag-NPs-TiO2 tiles, (i) dark run, (ii) photolysis, and (iii) photocatalytic experiments in the presence of Ag-NPs tiles under different light intensities were performed, and the results are shown in Figure 5. It is readily seen that no significant adsorption of E. coli on AgNPs-TiO2 tiles was found. Furthermore, the photolysis of E. coli without AgNPs-TiO2 tiles was not obvious within 300 min. However, in the presence of AgNPs-TiO2 tiles under light, a significant inactivation of E. coli that increases as a function of light intensity increases. This justifies the semiconductor behaviour of the Ag-species at the interface with bacteria [16,27,28]. Thus, these results indicate that the inactivation of E. coli is a light-dependent reaction owing to the formation of ROS on the surface of AgNPs-TiO2.

Different Initial Concentration of Bacteria
Photocatalytic experiments varying E. coli initial concentrations were conducted under low intensity solar simulated light (50 mW/cm 2 ) to further check the photocatalytic ability of Ag-NPs-TiO2 tiles ( Figure 6). It was found that, even at high concentrations of bacteria, the AgNPs-TiO2 ceramics are able to eradicate bacteria, hindering biofilm formation. In normal settings (hospital rooms, swimming pools, and so on), the concentration of bacteria is in the range of 10 2 -10 4 CFU/mL. The tiles inactivated these concentrations within only 60 min under low intensity solar light.

Different Initial Concentration of Bacteria
Photocatalytic experiments varying E. coli initial concentrations were conducted under low intensity solar simulated light (50 mW/cm 2 ) to further check the photocatalytic ability of Ag-NPs-TiO 2 tiles ( Figure 6). It was found that, even at high concentrations of bacteria, the AgNPs-TiO 2 ceramics are able to eradicate bacteria, hindering biofilm formation. In normal settings (hospital rooms, swimming pools, and so on), the concentration of bacteria is in the range of 10 2 -10 4 CFU/mL. The tiles inactivated these concentrations within only 60 min under low intensity solar light.  Figure 7 shows that the bacterial inactivation can take place under indoor light as well, which further confirms the good photocatalytic ability of such tiles. Under 5 mW/cm 2 , the photoactive tiles exhibited a 2.5-3 Log reduction of bacterial load. AgNPs-TiO2 tiles may be applied in indoor places   Figure 7 shows that the bacterial inactivation can take place under indoor light as well, which further confirms the good photocatalytic ability of such tiles. Under 5 mW/cm 2 , the photoactive tiles exhibited a 2.5-3 Log reduction of bacterial load. AgNPs-TiO 2 tiles may be applied in indoor places as anti-bacterial materials such as in kitchens and bathrooms, as bacteria are more likely to accumulate in indoor places.  Figure 7 shows that the bacterial inactivation can take place under indoor light as well, which further confirms the good photocatalytic ability of such tiles. Under 5 mW/cm 2 , the photoactive tiles exhibited a 2.5-3 Log reduction of bacterial load. AgNPs-TiO2 tiles may be applied in indoor places as anti-bacterial materials such as in kitchens and bathrooms, as bacteria are more likely to accumulate in indoor places.  Figure 8 depicts stereomicroscopy images captured on the surface of AgNPs-TiO2 tiles at different solar light irradiation times. The green dots represent the alive bacteria, while the red ones are dead bacteria. It can be seen that the green dots completely disappeared within 300 min of irradiation, while red dots appeared instead. The results presented in Figure 5 showed that total bacterial inactivation was reached within 180 min using the plate-count agar method. However, stereomicroscopy results show alive bacteria at this timing. These bacteria are viable non-cultivable as they were not able to develop colonies on the agar.   Figure 5 showed that total bacterial inactivation was reached within 180 min using the plate-count agar method. However, stereomicroscopy results show alive bacteria at this timing. These bacteria are viable non-cultivable as they were not able to develop colonies on the agar.  Figure 9 shows the micro-pH changes during the bacterial inactivation at the interface of the

Interfacial Potential and Micro-pH at the Tiles' Surface under Low Intensity Solar Simulated Light
(50 mW/cm 2 ) Figure 9 shows the micro-pH changes during the bacterial inactivation at the interface of the coated tiles. The initial surface potential was around 30 mV. When the light was switched on, the potential shifted to around 50 mV. The interfacial potential then decreased to reach 35 mV. This later phenomenon can be because of the (i) surface stabilization and equilibrium of the possible released ions, and (ii) accumulation of the bacterial biomass on the tile surface. This decrease is concomitant with the bacterial oxidation/destruction, as shown in Figure 5. The surface potential decrease is associated with the increase in cells' permeability, leaching out some ions, especially K + and Cl − . These changes of the surface potential were preceded by changes in pH. The initial pH of 7.2 decreases to 6.6 within the first 10 min. This can be because of the production of long-lived carboxylic acids intermediates during bacterial degradation. This slight decrease in pH could be also be because of the transformation of OH hydroxyls to • OH. HO are more likely to react with the positive holes to produce • OH than H 2 O molecules. Therefore, a slight decrease in pH could be detected during the photocatalytic experiments. Subsequently, the pH recovers its original level when the carboxylic acids decompose, generating CO 2 . A pH-shift between 6.6 and pH 7.1 takes place afterwards. This is equivalent to a fivefold increase in the concentrations of H + . The short carboxylic acids (branched or not) generated in solution during the bacterial inactivation period present a pKa~3. The increase of the pH again up to a pH~7 can be ascribed to the mineralization of the short carboxylic acids into CO 2 . This is the final step in the bacterial mineralization chain, as reported for many organic compounds and dyes like methylene blue. After total bacterial inactivation (300 min), the interfacial potential and the pH recover to stabilize at 50 mV and 7.1, respectively.  Figure 10 shows the E. coli inactivation on AgNPs-TiO2 ceramic tiles under low intensity solar simulated light (50 mW/cm 2 ) in the presence of • OH radical scavenger (DMSO), photogenerated hole (h + ) scavenger (EDTA-2Na), and O2 •− scavenger (SOD). It is readily seen from Figure 10 that the bacterial inactivation kinetics were not seriously affected when • OH radicals were scavenged. This suggests the weak contribution of • OH radical photo-generated at the interface of the AgNPs-TiO2 ceramic tiles. However, the bacterial inactivation kinetics were strongly inhibited when h + and O2 • were scavenged.  Figure 10 shows the E. coli inactivation on AgNPs-TiO 2 ceramic tiles under low intensity solar simulated light (50 mW/cm 2 ) in the presence of • OH radical scavenger (DMSO), photogenerated hole (h + ) scavenger (EDTA-2Na), and O 2

ROS Quenchers
•− scavenger (SOD). It is readily seen from Figure 10 that the bacterial inactivation kinetics were not seriously affected when • OH radicals were scavenged. This suggests the weak contribution of • OH radical photo-generated at the interface of the AgNPs-TiO 2 ceramic tiles. However, the bacterial inactivation kinetics were strongly inhibited when h + and O 2 • were scavenged.
(h + ) scavenger (EDTA-2Na), and O2 •− scavenger (SOD). It is readily seen from Figure 10 that the bacterial inactivation kinetics were not seriously affected when • OH radicals were scavenged. This suggests the weak contribution of • OH radical photo-generated at the interface of the AgNPs-TiO2 ceramic tiles. However, the bacterial inactivation kinetics were strongly inhibited when h + and O2 • were scavenged.   Figure 11. However, some reports suggested that the transfer of electrons from AgNPs to dissolved oxygen is hard and that electrons are most likely to move from excited AgNPs to CB of TiO 2 . Therefore, owing to the SPR of photoexcited AgNPs, more electrons are transferred from AgNPs to the conduction band of TiO 2 , which results in higher O 2 •− generation.
Unlike the photodegradation of organic compounds by ROSs, which is mostly based on the high oxidizing ability of species ( • OH is the most powerful), for the inactivation of bacteria, the long lifetime of oxidizing species is the most dominant factor rather than its oxidizing potential, because oxidizing species with longer lifetime can penetrate through the outer bacterial membrane and cell wall, resulting in its damage. Afterwards, the dimerization of OH produces the H2O2 species. Additionally, the H2O2 can also be produced via the double-h + oxidation of H2O. By analyzing Figure 10, we can notice that the scavenging of one of O2 •− or h + mostly limits the inactivation of E. coli, which means that the presence of only O2 •− or h + cannot really oxidize E. coli, and a synergetic effect as a result of the combination of O2 •− and h + is a must for efficient E. coli inactivation. As mentioned above, the oxidation of O2 •− with h + produces the oxiding species of 1 O2. However, as 1 O2 species, an excited state of O2, exhibit a very short lifetime (only 3 μs in water), the inactivation of E. coli by 1 O2 is not suggested. The synergistic production of H2O2 via two pathways, including the reduction of O2 •− and the dimerization of • OH, could be the reason behind the fast E. coli inactivation along with the effect of other ROSs. We suggest also that the presence of EDTA-2Na or SOD may prevent the reactivity of H2O2 for E. coli inactivation in this system. On the basis of this discussion, the plausible mechanistic pathways for the photocatalytic inactivation of E. coli on AgNPs-TiO2 ceramic tiles are shown in Figure 11. Photoproduced ROSs including • OH, O2 •− , h + , and H2O2 can directly inactivate E. coli. On top of that, AgNPs can play an important photonic and catalytic rule to visible light response of TiO2 and to decrease the recomination of electrons.  The stability of the photocatalytic activity was also investigated. The tile fabrics were washed with water after each photocatalytic test, dried at 80 • C, and a new cycle was started. The photocatalytic activity was seen to be stable over the ten tested cycles. The tiles exhibited similar photocatalytic bacterial inactivation, demonstrating the practical application of these scalable tiles.

Comparison of the Current Research and a Standard Test According to ISO22196:2011
The results of the present study were finally compared to the antimicrobial activity of the AgNPs-TiO 2 photocatalytic ceramic tile and a traditional commercial tile based on the ISO (International Standards Organization) 22196 method, the official standard for measuring the antibacterial efficiency of advanced ceramics [31]. In particular, the main goal was the evaluation of the antibacterial efficiency of the two materials under dark conditions, exclusively exploiting the presence of Ag and not of the photocatalyst.
All specimens (50 × 50 mm 2 ) were rinsed with distilled water and autoclaved at 121 • C for 30 min before the analysis in order to remove any organic residue on the surfaces. For each strain, six tile samples without (controls) and six with photocatalytic coating were investigated for 24 h (time required by the norm), at 37 • C, with an initial inoculated volume of 150 µL E. coli (ATCC 8739) and a concentration in the range 10 5 cells/mL, one/two order of magnitude less than that used in our research in Figure 5. The results are reported in Table 2.
As expected, on the traditional tile, the number of bacteria increased after 24 h owing to the temperature and environment test conditions that facilitate the bacteria growth. The tile doped with AgNPs-TiO 2 shows a drastic decrease of the bacteria number with a complete bacteria inactivation after 24 h.

Conclusions
Sustainable plasmonic AgNPs-TiO 2 digitally printed onto commercial porcelain-grès tiles showed an excellent continuous photocatalytic antibacterial activity under indoor and outdoor light irradiations. A homogenous surface and a great stability were obtained because of the new economical digital printing synthesis method used for the deposition of AgNPs-TiO 2 onto the surface ceramic tiles. A synergistic effect resulting from the combination of charges separation and SPR effect at the interface of AgNPs and TiO 2 leads to enhancing the overall photoactivity of the AgNPs-TiO 2 for generating more ROS species for bacterial degradation. E. coli at a concentration ranging from 10 4 to 10 7 CFU/mL was totally inactivated under low intensity solar simulated light (50 mW/cm 2 ). ROS quenching experiments confirmed that the formation of ROSs is the reason behind the bacterial degradation under light irradiation.
Tests in the dark confirmed the role of Ag in the bacteria inactivation even in the absence of light, depending on the number of bacteria present at the ceramic surface. Even at high concentrations of bacteria, the AgNPs-TiO 2 tiles are able to eradicate bacteria, hindering biofilm formation.