Biocompatibility and antibacterial properties of TiCu(Ag) thin films produced by 2 physical vapor deposition magnetron sputtering

26 Mechanical robustness, biocompatibility, and antibacterial performance are key features 27 for materials suitable to be used in tissue engineering applications. In this work, we 28 investigated the link existing between structural and functional properties of TiCu(Ag) thin 29 films deposited by physical vapor deposition magnetron sputtering (MS-PVD) on Si 30 substrates. Thin films were characterized by X-ray diffraction (XRD), nanoindentation, 31 atomic force microscopy (AFM), and X-ray photoelectron spectroscopy (XPS). The TiCu(Ag) 32 films showed complete amorphous structure and improved mechanical properties in 33 comparison with pure Ti films. However, for contents in excess of 20% Ag we observed the 34 appearance of nanometric Ag crystallite. The TiCu(Ag) thin films displayed excellent 35 biocompatibility properties, allowing adhesion and proliferation of the human fibroblasts 36 MRC-5 cell line. Moreover, all the investigated TiCu(Ag) alloy display bactericidal 37 properties, preventing the growth of both Pseudomonas aeruginosa and Staphylococcus 38 aureus . Results obtained from biological tests have been correlated to the surface structure 39 and microstructure of films. The excellent biocompatibility and bactericidal properties of these 40 multifunctional thin films opens to their use in tissue engineering applications. 41 42 43


Introduction
The aging of world population drives an increasing demand of tissue and organ replacements [1].To date, more than 10 million transplantations are performed annually, with an yearly increase of about 6%, and an overall cost of more than $500 billion per year [2].
However, tissue and organ transplantations present two major limitations: the low availability of donors and/or risk of disease transmission and immune rejection [3][4][5].
Tissue engineering (TE) is an emerging and promising alternative approach of biomedicine to treat or to replace damaged tissues and organs.TE combines materials science, chemistry, physics, and cell biology to allow tissue and organ repair or reconstruction.TE is often based on nanoscaffolds enabling cell adhesion, migration, proliferation, and differentiation [6][7][8][9].The properties of a scaffold mainly depend upon the types of biomaterial and fabrication techniques [8].In particular, the size, the shape, and patterning of adhesion sites are crucial elements in the design of effective scaffold surfaces.Nano-scaffolds must be biocompatible and can be combined with organic and inorganic materials to mimic the structure and function of the natural extracellular matrix (ECM).The ECM allows the cells to accomplish the biochemical and biophysical functions related to tissue and/or organs regeneration [8,10].Cells grown on nanoscaffolds can generate biocompatible, immunocompatible, and biofunctional tissues inside the body, counteracting the drawbacks associated with autologous grafting and allograft tissue transplantation, thereby alleviating the risk of rejection [8][9].
Another critical issue of nanoscaffolds is their capability to prevent microbial growth.
Antimicrobial capability is conventionally obtained by means of biochemical approaches relying on nanoscaffold coating with biocidal substances such as silver and antibiotics.
In this complex framework, surface engineering and the development of nanostructured thin films is gaining importance, especially for applications where a combination among surface hardness (or wear resistance), biocompatibility, and antibacterial performance are desired.Recently, multi-element thin films, such as Zr-based thin films (e.g., Zr-Cu, Zr-Cu-Ag, ZrCN, Zr/ZrCN multilayer) [12,13] and Ti-based thin films (e.g., TiN, TiCu, Ti-Zr-Si) have emerged as a new class of nano-engineered thin films, featuring an excellent combination of high mechanical strength and biocompatibility.Moreover, these films are promising systems for biocompatible coating deposition.Indeed, the use of physical vapor deposition (PVD) for their growth allows a fine control of the material nanostructure, which leads to increased hardness and wear resistance [14,15].The desired combination of mechanical strength, biocompatibility, and antimicrobial activity can also be achieved by constructing multi-layers.Recent studies have confirmed the potential antibacterial behavior of Au, Cu, Zn, Ag additions to Ti-based films [16,17].The biocidal performance of Cu is linked to the release of Cu +1 and Cu +2 ions, as observed in TiCu [18,19].Recently, Cu-based systems have also been proposed engineer surfaces with antiviral properties, also in the framework of the COVID-19 pandemic [20].Very recent examples include Cu-coated touch surface fabricated by cold-spray technology, as well as antiviral CuxO/TiO2 photo catalyst thin films with photo-activated anti-viral properties [21,22].Consequently, the TiCu systems are of particular interest for the generation of material systems, featuring both anti-infective properties and surface hardness, as required in human implants and/or touch surfaces [23,24].
In this context, the development of antibacterial metallic thin films combining biocompatibility with relatively high surface hardness, as witnessed by the growth of human fibroblasts, is here reported.To this aim, in this work we investigate TiCu(Ag) PVD sputtered thin films to evaluate their potential as biomedical thin films.By using high-resolution surface chemical and morphological characterization combined with cell growth studies and antibacterial tests, significant biocompatibility and antibacterial properties have been observed, and correlated to film structure and surface properties.

Thin film deposition
TiCu(Ag) thin films were deposited on 217 mm 2 coupons extracted from 4" Si(001) undoped wafers by means of direct current (DC) magnetron sputtering in a deposition system equipped with three unbalanced magnetrons.For the deposition, we have used Ti, Cu, and Ag targets with a 3" diameter and 99.99% purity were employed.The Si substrates were cleaned in ultrasonic bath and ethanol for 10 minutes before mounting them on the substrate holder.
An Ar + sputtering step (powered by radio frequency (RF) power supply at 50 KHz, at the Ar pressure of 1.2 Pa and a discharge power of 0.03 KW) was performed to clean and activate the Si surface immediately prior to metal deposition.The distance between the substrates and targets was 70 mm, while the substrate was kept in rotation at 80 rpm.All the depositions were performed at a 0.52 Pa Ar pressure (chamber base vacuum of 1.0×10 -5 Pa), with no intentional substrate heating.By applying different DC-power to the targets for 40 min, 4 sample sets were obtained, always keeping the Ti:Cu ratio equal to ~1.Deposition conditions, thickness, and composition are listed in Table 1.

Characterization of thin films
Crystallographic structure of the thin films was carried out by X-ray diffraction (XRD), using a θ-2θ Bruker D8 Advanced system with Cu Kα radiation (λ = 0.154 nm).Diffraction scans were performed by using grazing incident angle of 0.75 degree with time step of 0.02˚/sec.The chemical composition of the thin films was estimated via energy dispersive Xray spectroscopy (EDX, Oxford instrument INCA), using built-in sensitivity factors for calibration.The film thickness was measured by using a white light optical profilometer with a Leica DCM 3D software package via automatic step measurement of the coated and the uncoated parts of the substrate.
The elastic modulus (E) and hardness (H) values were determined using nanoindentation testing method using a KLA-Nanomechanics G200 fitted with a Berkovich diamond indenter operating in continuous stiffness measurement mode, hence allowing obtaining both E and H as a continuous function of the depth from a single indentation experiment [25].A standard fused silica sample was tested before and after a batch of measurements to calibrate the tip, so to ensure the reliability of the results.A least 25 indentations were performed on each sample.Calculations were made by the Oliver and Pharr method from the load-displacement curve using 10% of the film thickness at the maximum indentation depth [26].

X-ray Photoelectron Spectroscopy (SR-XPS)
X-ray Photoelectron Spectroscopy (SR-XPS) measurements were performed at the materials science beamline (MSB) of the Elettra synchrotron radiation source (Trieste, Italy).
The UHV end station, with a base pressure of 2×10 −10 mbar, is equipped with a SPECS PHOIBOS 150 hemispherical electron analyzer and a dual-anode Mg/Al X-ray source, an ion gun, and a sample manipulator with a K-type thermocouple attached to the rear side of the Calibration of the energy scale was made referencing the spectra to the C1s core level signal of aliphatic C atoms (285.0 eV).Curve-fitting analysis of the experimental spectra was carried out using Gaussian curves as fitting functions.The Ti2p3/2,1/2 core level were fitted using a spin−orbit splitting of 5.7 eV and a branching ratio (2p3/2/2p1/2) of 2; the Cu2p3/2,1/2 doublets were fitted using a spin-orbit splitting of 19.8 eV and a branching ratio (2p3/2/2p1/2) of 2. For the Ag3d5/2,3/2 doublets, a splitting of 6.0 eV and a branch 3d5/2/3d3/2 ratio of 3/2 were used.When different species were identified in a spectrum, the same Full Width at Half Maximum (FWHM) value was set for all individual photoemission peaks.Atomic ratios were calculated from peak intensities by using Scofield's cross section values.

Surface sterilization
Thin films were rinsed in 70% ethanol in sterile deionized water and then flamed with a Bunsen burner.Sterilization was performed under biosafety cabinets with installed HEPA filters to avoid contamination.After sterilization, films were air-dried, and structural and mechanical properties were evaluated in order to verify stability prior to testing for biocompatibility and antibacterial properties.

Evaluation of surfaces biocompatibility
To correlate MRC-5 cell density to the relative luminescence units (RLU) value, a calibration curve was set up.With this aim, MRC-5 cells were plated in triplicate in opaque 96-well plates at a density of 2.5×10 3 , 5.0×10 3 , 1.0×10 4 , 2.0×10 4 , and 4.0×10 4 cells/well.On the basis of the results obtained from the calibration curve, 1.0×10 4 MRC-5 fibroblasts were seeded on sterile TiCu surfaces coated with 0%, 10%, 20%, and 30% Ag thin films placed into opaque 96-well plates.Cell growth and proliferation was assessed by incubating MRC-5 fibroblasts for 10 min with the CellTiter-Glo® Luminescent reagent (Promega, Madison, WI, USA) added in a 1:1 ratio with the complete cell culture medium.Luminescence was measured using the Tecan Spark 10M plate reader (Tecan, Männedorf, Switzerland).Background luminescence was measured in the complete culture medium without cells, and subtracted from each experimental value.

Testing of the antibacterial properties
Bacteria were routinely grown in Nutrient Broth (NB) No. 2 (# CM0067B; Thermo Scientific™, Waltham, Massachusetts, USA).The day before the experiment, glycerol stocks of S. aureus ATCC 25923 or P. aeruginosa ATCC 15692 (strain PAO1) were streaked on NB supplemented with 15% agar (NA) plates and incubated at 37 °C for 24 h.By using sterile inoculating loops, bacterial colonies were transferred in 1 mL NB diluted 1:500 (NB1:500) in deionized sterile water and the bacterial concentrations was adjusted to ~5.0×10 7 colony forming unit (CFU)/mL.surface, as well as the glass control surface, was placed into a Petri dish (⌀ 3 cm).Then, 0.005 mL of a suspension of either S. aureus or P. aeruginosa at a concentration of ~ 5.0×10 7 colony forming unit (CFU)/mL was dripped onto the surfaces and samples were incubated at 37 °C overnight (ON) at 99% relative humidity.After incubation, each surface was placed into 1 mL NB1:500 at room temperature for 15 min, and then vortexed for 1 min to allow the detachment of bacteria from the surface.The bacterial suspension was appropriately diluted and plated on NA for CFU counts.To determine the CFU at time 0 h, suspensions of either S. aureus or P.
aeruginosa (presumptive concentration ~5.0×10 7 CFU/mL) were appropriately diluted in saline and plated onto NA.At least two samples were assessed for each bacterial strain.The antibacterial activity (BA) was calculated by the following formula: Where, N0 h and N24 h are the CFU average numbers counted at 0 h and 24 h, respectively, for each type of surface.
An agar diffusion assay was performed to detect the release of bacterial growth inhibitors.A suspension of either S. aureus or P. aeruginosa (OD600 = 0.1) was uniformly spread onto NA plates using a sterile cotton swab.TiCu and TiCu(Ag) surfaces were placed on the NA inoculated plates.The glass surface was used as negative control, whereas antibiotic discs (i.e., erythromycin, E 15 µg; and amikacin, AK 30 µg) were used as positive control of bacterial inhibition.After 16-h incubation at 37 °C, the release of antibacterial factors by TiCu(Ag), glass surfaces, and antibiotic discs was visually assessed by the presence of the inhibition zone around the sample.

Morphological characterization
The surfaces were morphologically characterized by AFM and optical microscopy; the results obtained were compared to control samples represented by a suspension of bacteria poured directly on the TiCu surface.AFM measurements were performed using a Dimension ICON AFM (Bruker, Santa Barbara, CA) operating in peak-force mode.The AFM was equipped with a ScanAsyst-Air Bruker silicon probe featuring a nominal cantilever elastic constant of 0.4 N m −1 and a tip with a nominal radius of 2 nm.The oscillation frequency and oscillation amplitude of the cantilever were set to 1 kHz and 150 nm, respectively.For each measurement, height sensor and peak force error images were recorded simultaneously.The AFM images were analyzed and processed with the software Gwyddion [28], applying a firstorder flattening.Surface roughness was obtained by measuring the root-mean-square deviation of surface heights on 20×20 μm 2 images.Optical microscopy images were acquired using a Nikon Eclipse ME600 microscope equipped with Nikon DXM1200 digital camera (Nikon, Tokio, Japan).

Structural, morphological, chemical, and mechanical characterization of substrates
The X-ray diffraction patterns of the as-deposited thin films (see Table 1) are shown in No impact on the XRD spectra has been observed flaming with alcohol. 13 The AFM analysis revealed crack-free smooth surfaces for all thin films with an increase in the average roughness as the Ag content increased from 10 to 30%.The amorphous TiCu and TiCu(Ag) thin films with the Ag content below 20% exhibited an average surface roughness well below 0.5 nm.For TiCu(Ag) film featuring Ag content > 20%, we observed the appearance of nanocrystallites, which, in agreement with the XRD evidence above discussed, was attributed to the formation og Ag precipitates.Consequently, the surface roughness increases up to ~ 1 nm in the Ag-richest sample (see Fig. 3).It is worth noting that the surface clusters observed by AFM have a size larger than that obtained by the XRD analysis.This is not in contrast with our hypothesis, since the surface cluster might be formed by the assembly of different nanocrystallites of a lesser size.Thin films with such a low average surface roughness are in general very favorable for antibacterial and biomedical applications [31], especially if antimicrobial agents such as Cu and Ag are added into the protective thin film, which induce a release of metallic ions after exposure to a humid environment [32].The elastic modulus E and hardness H of thin films were calculated as a function of Ag contents by the nanoindentation method.The TiCu film exhibited the highest E = 124.3GPa and H = 7.83 GPa values.As we can observe in Fig. 3 (left axis), the addition of Ag into the TiCu(Ag) thin films induced a decrease of both modulus and hardness, with the lowest values of modulus (109.25 GPa) and hardness (6.45 GPa) observed in the sample containing 30% of Ag.
After the flame sterilization test, no change was observed in the mechanical properties of thin film, consistent with a previous report [33].All the samples were analyzed by XPS spectroscopy before and after flaming with alcohol (flamed samples will be labelled (F) in the following text).The measured binding energies (BE, ±0.2 eV), Full Width at Half Maximum, and atomic ratios calculated from peak areas for all the analyzed samples are reported in Table S1.
In Fig. 4 is reported as an example the Ti2p, Cu2p, and Ag3d spectra and the relative curve-fitting analysis for the sample TiCu-30%Ag (F).The measured BE value of the Ti2p3/2 signal (458.7 eV) corresponds to the expected value for TiO2 [34]: indeed, when exposed to air, Ti is always oxidized to titania in the outmost surface layer [35].The Cu2p3/2 signal results from two components peaks located at 933.0 eV and 935.0 eV that have been attributed to metallic (Cu) and oxidized (CuO) copper, respectively.Moreover, the presence of a shake-up satellite, evident in the spectrum at about 943.5 eV, is a distinctive feature of Cu in the +2 oxidation state [36].Finally, the Ag3d5/2 main peak position at 368.5 eV is typical of metallic Ag; a small higher BE component, about 10% of the main component peak, located at about 369.9 eV can also be observed in the spectrum, and can be attributed to oxidized, positively charged silver atoms, indicated as Ag2O in Table S1 [37].In all samples, Ti was completely oxidized to Ti(IV), Cu was partially oxidized to Cu(II), and most Ag was predominantly in the unoxidized metallic state, consistent with the different reactivity towards oxygen of the three metals (Table S1).Peak areas have been used to calculate atomic ratios between the oxidized and metallic components and between the three elements present on the sample surface (Table S1).The measured atomic ratios between copper and titanium disagree with the expected value of 1:1, possibly because XPS is a surface-sensitive technique with a sampling depth of approximately ~5 nm.Therefore, the difference between measured and expected elemental composition is relative only to the outmost sample surface.
For TiCu samples, the copper concentration on the sample surface was lower than expected.On the other hand, the Ag:Ti atomic ratio on the surfaces of TiCu(Ag) thin films was higher than those of the targeted bulk values; a slight saturation effect has been observed at 30% Ag concentration.Apparently, when Ag was introduced in the mixture, Cu and Ag form an alloy and migrate together to the outmost sample surface, with Ag forming clusters as evidenced by AFM and XRD analysis (see Figures. 1 and 2).
We notice that the flaming procedure does not affect the sample surface composition and the oxidation state of the three elements.

Biocompatibility studies of the TiCu(Ag) surfaces
The TiCu(Ag) thin films biocompatibility has been evaluated by measuring human MRC-5 fibroblasts viability based on ATP production under aerobic conditions, which reflects the presence of metabolically active cells (Figure 5A).With this purpose, we first determined the correlation between the number of viable cells and the emitted luminescence (RLU) (Figure 5B).On the bases of the results obtained, we seeded 1.0×10 4 MRC-5 cells on polystyrene cell culture plate (Ctrl) and on sterile TiCu surfaces coated with 0%, 10%, 20%, and 30% Ag has been evaluated (Figure 5C).As shown, the number of cells after 24 h from seeding was comparable between Ctrl, TiCu, and TiCu(Ag) thin films, thus indicating that all the tested surfaces did not affect cell viability and proliferation.Next, in the perspective of using these surfaces for TE studies, we evaluated if the trypsin-mediated detachment of cells affected the biocompatibility of TiCu and TiCu(Ag) thin films.Results obtained indicated that trypsinization did not alter the capability of the TiCu(Ag) thin films to allow MRC-5 growth, supporting their possible use for tissue culture studies (Figure 5D).

XPS analysis of TiCu(Ag) surfaces after human MRC-5 cells trypsinization
To evaluate if thin film could be re-used for biocompatibility test after human cells detachment by trypsinization, TiCu(Ag) surfaces were analyzed by XPS and C1s, N1s, O1s, Ag3d, Ti2p and Cu2p core levels were investigated.XPS spectra and data (BE, FWHM, and atomic ratios) are reported in Figure 6 and Table S2.As already evidenced for the pristine samples, the Ti2p spectra are typical of TiO2.The Cu2p spectra revealed the presence of both metallic and oxidized (CuO) copper, while the Ag3d signal is typical of metallic silver with a very small component related to oxidized silver (spectra not shown).No relevant change were evidenced on the oxidation state of metals on the sample surface.C1s, O1s and N1s spectra revealed the presence of organic molecules, particularly peptides, deposited on the TiCu(Ag) film surfaces.assigned to the oxygens of titania; (ii) the peak at about 532.0 eV, assigned to O=C oxygens of the peptide backbone; (iii) the peak at 533.5 eV assigned to C-O oxygens; and (iv) the peak at nearly 535.0 eV, related to physiosorbed water [32].
In summary, the XPS data analysis points to the presence peptide residues on the samples after human cells detachment, possibly representing cells residues of previous MRC-5 cell growth, not completely removed from the thin films by trypsin, and/or trypsin residues adsorbed on the sample surface.However, the presence of these peptode residues did not influence MRC-5 adhesion and proliferation, as reported in Figure 5D.

Evaluation of the bactericidal properties of the TiCu(Ag) surfaces
The bactericidal property of TiCu and TiCu(Ag) thin films was evaluated by testing the growth of two well-known nosocomial pathogens, i.e. S. aureus and P. aeruginosa.The experimental protocol is illustrated in Figure 7A.Briefly, S. aureus and P. aeruginosa were dispersed on either a glass surface or TiCu thin films coated with 0%, 10%, 20%, and 30% Ag. Bacteria were incubated for 24 h at 37°C in a controlled (99%) humidity chamber prior to mechanical detachment and cell (CFU) counting.For both species, limited or no loss of bacterial viability was observed on glass after 24 h (Figure 7B).Conversely, no colony growth was observed when bacteria were dispersed on TiCu and TiCu(Ag) surfaces, indicating a strong bactericidal activity (BA), even in the absence of Ag (Figure 7B).To evaluate the release of antibacterial agent(s) from the TiCu(Ag) surface, a suspension of S. aureus or P.
aeruginosa was spread over NA plates, then TiCu(Ag) surfaces were placed onto the inoculated plates.The glass surface was used as the negative control (no growth inhibition), whereas antibiotic discs (i.e., erythromycin, E 15 µg, and amikacin, AK/30 µg) were used as the positive control (growth inhibition due to antibiotic diffusion around the discs).No growth inhibition was observed around both glass and TiCu(Ag) surfaces, indicating no release of inhibitory agent(s), as opposed to the large inhibition halo around the antibiotic discs (Figure 7C).Antibiotic discs and glass were used as positive and negative controls of bacterial inhibition.
To exclude incomplete detachment of S. aureus and P. aeruginosa from NB1:500-washed TiCu surfaces, both optical microscopy (Figure 8A) and AFM analyses (Figure 8B) were performed.Microscopy results were compared with positive control (Ctrl) samples in which the same number of bacteria was directly dispensed onto the TiCu surfaces, without subsequent washing step.A complete detachment of bacterial cells from TiCu surfaces was observed for both strains.Indeed, the large-scale optical microscopy images showed that on Ctrl samples many "coffee-stain" clusters were present (Figure 8A), while no clusters were observed on washed TiCu surfaces.This was confirmed by acquiring several images across the surfaces.By AFM characterization, the clusters on Ctrl samples were univocally identified as bacterial aggregates (Figure 8B).Notably, S. aureus bacteria showed a round shape, with an average diameter and height of 950 nm and 530 nm, respectively.The P. aeruginosa bacteria were instead shallower with an average height of 230 nm and showed an elongated shape, often with flagella, having an average long-axis length of 1300 nm.The tendency of bacteria towards clustering in Ctrl samples, evident in Figure 7B, was observed for similar deposition conditions and attributed to the capillary flow induced by the evaporation of the drop deposited [39,40].Conversely, the TiCu surfaces analyzed after bacteria detaching showed morphological features with heights below 30 nm (thus more than one order of magnitude lower than typical bacteria heights), likely to be attributed to residual of the nutrient medium used for the growth.The rms surface roughness evaluated on 20×20 μm 2 images was 1.71 nm and 1.87 nm after detachment of S. aureus and P. aeruginosa, respectively; while it exceeded 50 nm for the Ctrl surfaces with bacteria adhered.This comparative morphological analysis confirms that: (i) bacteria were completely detached with the experimental protocol used for the plate-count experiments and, (ii) TiCu surfaces were bactericidal for both S. aureus and P. aeruginosa.Images were acquired after bacteria detaching and were compared to samples (Ctrl) where a suspension of each bacteria strain, containing the same number of bacteria as in the experiment, was poured directly on the TiCu surface.In the insets shown in panels B, 3D zoom ups are displayed.Here, the biocompatibility and the antibacterial properties of TiCu and TiCu(Ag) thin films, as produced by PVD magnetron sputtering method, are reported.All TiCu and TiCu(Ag) thin films display a good adhesion and growth of human fibroblast MRC5 cells, together with strong antibacterial activity against both S. aureus and P. aeruginosa, here considered as prototypes of nosocomial bacterial pathogens.Results are extremely significant for TE applications, biocompatibility and antibacterial performance representing key design parameters for biomaterials.In fact, tissue and implant-associated bacterial infection is a growing problem, responsible for increased morbidity and mortality, together with enormous economic losses to the public health system.Bacteria can tightly adhere to the biomaterial surface, and the formation of biofilm can help bacteria to escape the host immune system and antibiotics.In turn, this causes the emergence of bacterial resistance to antibacterial drugs and finally determines implantation failures [41].Therefore, implanting materials that combine the capability to favor eukaryotic cells adhesion and proliferation for tissue regeneration with strong antibacterial properties are urgently needed.Metallic elements (e.g., Au, Ag, Cu, and Zn) have been proven to exert antibacterial activity by surfaces coating or doping [42].

Discussion
Remarkably, coating displays an amorphous single-phase structure with improved mechanical properties with respect to pure titanium up to Ag ~20%.
A main discussion point on the observed results should be the interpretation of surface chemical characterization of the films in relation with the observed functional properties.In fact, XPS analysis shows that TiO2 represents the main oxidation state for titanium, thus explaining the observed biocompatibility.At the same time, all films show the presence of a combination between metallic and oxidized copper, with an increase of Cu/Ti ratio after addition of Ag.In addition, Ag is shown to be predominantly present in the metallic state with low levels of Ag2O on the surface.Such observations can be used to explain the relevant antibacterial performance of the films.In fact, the presence of a single-phase film with homogeneous surface distributions of Ag and Cu can be very effective in protecting the surface against bacteria, while maintaining a high biocompatibility thanks to the concurrent homogenous distribution of TiO2.
The biocompatibility of TiCu(Ag) thin films here ported is in line with the very low cytotoxicity observed in Ag when tested in human blood, adenocarcinomic alveolar basal epithelial cells, liver cancer cells, breast cancer cells, and gastric cancer cells [43][44][45].
Noteworthy, several data support the anti-bacterial, anti-viral, anti-biofilm, and antiinflammation activity of Ag, especially at nanoscale [43][44][45][46][47]. Furthermore, Cu is a transition metal and an essential micronutrient in humans and bacteria.Indeed, Cu is involved in many biosynthetic and metabolic processes, being a cofactor of many redox enzymes and playing a role in iron metabolism.In humans, Cu plays also an essential role in immune function [48].
It has been reported that TC50 values (i.e., the value that indicates which is concentration of a specific substance that produces toxic effects in 50% of the cell population) for Cu is 344±4.4 M in human gingival fibroblast [49], which means that it exerts a very low cytotoxicity.
Besides, while low concentrations of Cu are essential for bacteria metabolism, high concentrations, cause cell growth inhibition or even cells death [50,51].Therefore, Cu represent an optimal metal to prepare antibacterial titanium alloys [42,52,53].

Conclusion
In this paper, the surface structural and functional properties of amorphous TiCu(Ag) thin film have been investigated.The increase of Ag content is accompanied by the appearance of Ag-nanocrystallites and by a decrease of both elastic modulus and hardness of the thin film.Based on the multi-technique characterization and cellular studies, it can be concluded that binary TiCu and TiCu-10% Ag showed the best mechanical properties with amorphous glassy structure combined to excellent biocompatibility and antibacterial activity.On the contrary, ternary TiCu(Ag) thin films with 20% Ag content showed moderate mechanical properties, although they display excellent biocompatibility and antibacterial properties.
The excellent biocompatibility and bactericidal properties of these multifunctional thin films opens to their use in TE applications.Moreover, these types of thin films could be also used to coat surgical tools and hospital furnishing.For the future, it will be necessary to test these thin films for the biocompatibility of mesenchimal stem cells, one of the most studied stem cells for their great potential to enhance TE thanks to their capacity to differentiate into cartilage, bone, fat, muscle, tendon, skin as well as hematopoietic-supporting stroma and neural tissue [54].Moreover, these kinds of thin films could be suitable for applications also against SARS-CoV-2 infection [20].Indeed, Cu and its oxide have been demonstrated to act as efficient antiviral agents [20,21,55].

Figure 1 .Figure 1 .Figure 2 .
Figure 1.The diffraction pattern of the TiCu(Ag) samples shows only a broad band in the [38-

Figure 4 .
Figure 4. Comparison of elastic modulus and surface average roughness (A) and of hardness

Figure 5 .
Figure 5. Evaluation of the biocompatibility of the TiCu and TiCu(Ag) surfaces.(A)