Fabrication and Characterization of Antimicrobial Magnetron Cosputtered TiO 2 /Ag/Cu Composite Coatings

: The aim of this study was to prepare TiO 2 /Ag/Cu magnetron co-sputtered coatings with controlled characteristics and to correlate them with the antimicrobial activity of the coated glass samples. The elemental composition and distribution, surface morphology, wettability, surface energy and its component were estimated as the surface characteristics inﬂuencing the bioadhesion. Well expressed, speciﬁc, Ag/Cu concentration-dependent antimicrobial activity in vitro was demonstrated toward Gram-negative and Gram-positive standard test bacterial strains both by diffusion 21 assay and by Most Probable Number of surviving cells. Direct contact and eluted silver/coper nanoparticles killing were experimentally demonstrated as a mode of the antimicrobial action of the studied TiO 2 /Ag/Cu thin composite coatings. It is expected that they would ensure a broad spectrum bactericidal activity during the indwelling of the coated medical devices and for at least 12 h after that, with the supposition that the beneﬁts will be over a longer time.


Introduction
The increasing number of medical devices-associated infections and the microbial resistance to conventional drugs treatments raised the need for the efficient antimicrobial protection of medical devises, which is especially important for those in the urinary tract [1]. Surface coating technologies are a tool to mitigate the problem. They usually aim at the creation of an antiadhesive surface that does not allow the attachment of microbial cells and hence inhibits the initial stage of the biofilm formation; or at the creation of a contact killing or antibacterial agent delivering surface. Antibiotics, antimicrobial peptides and other organic compounds, such as enzymes, polymer gels, carbon materials, metal, metal oxide nanoparticles and ions and others, are in use in the development of antimicrobial coatings [2,3].
Due to their advantages, like low-toxicity, biocompatibility, high stability, etc. functional nanostructured TiO 2 coatings are one of the most investigated among the metal oxide coatings [4][5][6][7]. Doping with other metal or/and metal oxide nanoparticles (NPs) to enhance their antimicrobial activity is widely investigated. Ag and Cu (which have been known for centuries for their good antibacterial properties) are one of the most often used. According to some current reports, Cu is also active against viruses [8][9][10][11]. The quantitative control over the incorporated Ag and Cu amount allows the realizing of both antibacterial activity and biocompatibility [12,13].
Our short review demonstrated that different techniques are in use for preparation and deposition of antimicrobial coatings, each one with its own advantages and disadvantages. Among them are the Chemical Vapor Deposition (CVD) in different modes [18][19][20][21][22][23][24]; the solgel method [41][42][43]; the sputtering of multilayered and nanocomposite thin films [3,[44][45][46][47][48][49]; and others. Magnetron sputtering is preferred for the fabrication of antimicrobial coatings because of its advantages. This is a dry method that ensures precise control over the plasma processing parameters and allows good control over the thickness of the deposited layers, their chemistry, charge, structure, morphology, hydrophilic/hydrophobic balance, etc. The long-term antimicrobial activity of the magnetron sputtered thin films, which could be deposited on any surface, is another advantage [48].
Triple TiO 2 /Ag/Cu magnetron cosputtered thin composite coatings are scarcely studied although they hold promise for a broad spectrum antibacterial activities due to the combination of three types biologically active nanoparticles: TiO 2 , Ag and Cu. In a former paper [50] we reported some preliminary results about the bioactivity of such coatings without correlation to their characteristics. Therefore, the fabrication of and search for a correlation between the biological activity and the characteristics of magnetron cosputtered TiO 2 /Ag/Cu composite coatings became the subject of this investigation, with our final aim being to demonstrate the ability of such coatings to be adjusted to the antimicrobial protection requirements of different medical devices.

Preparation of Coated Samples
Alcatel DION 450 magnetron installation, equipped with rotary and diffusion vacuum pumps and cylindrical glass camera was used to deposit thin coatings (50-60 nm, Tailor Hobzon prophilometer, Leicester, UK) on glass substrates (20 mm × 20 mm × 1 mm). RF magnetron cosputtering was performed under the following operation conditions: power of 50 W; vacuum in the sputtering chamber of 10 −3 Pa before the film deposition; 0.8 Pa argon gas atmosphere; distance between the target and substrate of 7.5 cm; Agand Cu-plates, placed together with TiO 2 target in its erosion zone; without heating. The glass substrates were 76 preliminary treated with Piranha etch, H 2 SO 4 :H 2 O 2 (30%) = 3:1, to improve their adhesion with the coatings. The area of TiO 2 was of 750 mm 2 , whereas that of both Ag and Cu sheds was varied between 40 and 190 mm 2 to obtain thin coatings with different Ag and Cu content was confirmed by Energy Dispersive X-ray Detector (EDX) analysis. Five magnetron cosputtered TiO 2 /Ag/Cu coatings, deposited on glass substrates were selected for this investigation to evaluate the influence of the Ag and Cu content on their characteristics and antibacterial activity: three of them with varied Cu content at almost equal Ag amount (Samples 1, 2 and 3) as well as three of them with varied Ag content at almost equal Cu amount (Samples 3, 4 and 5). The exact content of Ti, Ag and Cu, at.%, was evaluated from the detailed Ti2p, Ag3d and Cu2p X-ray Photoelectron Spectroscopy (XPS) spectra of the corresponding coating.

X-ray Photoelectron Spectroscopy (XPS)
Kratos AXIS Supra spectrometer (Manchester, UK) with a monochromatic Al X-ray source under vacuum better than 10 −8 Pa at 90 • take-off angle was used for elemental analysis of the coatings. It started with a survey scan from 0 to 1200 eV, pass energy of 160 eV at steps of 0.5 eV with 1 sweep. For high resolution analyses the sweep number was increased, the pass energy was lowered to 20 eV at steps of 100 meV. The C1s photoelectron line at 285 eV was used for calibration of the spectra. Ag content, atomic %, was calculated from the area of the detailed Ag 3d3/2 and Ag 3d5/2 peaks; that of Cu-from the area of Cu2p3/2 and Cu2p1/2 peaks. SEM (JEOL, model JSM-35 CF, Tokyo, Japan) apparatus was used to observe the morphological features of the studied magnetron cosputtered TiO 2 /Ag/Cu composite coatings as well as of E. coli cells damaging and contact killing by them. The samples were gold sputtered coated and viewed in the second electron mode with field emission gun.
Lira/TEScan SEM (Brucker, Berlin, Germany) apparatus equipped with Quantax 200 EDX detector was used for preliminary evaluation of the elemental composition and to observe elements' distribution in the studied magnetron cosputtered TiO 2 /Ag/Cu thin composite coating. Every test culture was grown in a nutrient broth (Conda, Torrejon de Ardoz, Spain) on a rotary shaker (180 rpm) at 37 • C overnight followed by a subculturing of the bacteria in fresh medium until optical density reached 0.02-0.05, detected spectrophotometrically at 610 nm (0.5 McFarland). The bacterial suspension (100 µL) was used for inoculation of nutrient agar in Petri dishes.

Agar Diffusion Assay
Agar diffusion assay was used for a fast antibacterial activity screening of the studied TiO 2 /Ag/Cu coatings. The aliquots of 100 µL microbial suspension were randomly spread on a solid nutrient medium (Conda, Spain). The coated samples were sterilized by UVirradiation for 30 min and placed on inoculated agar. The inoculated dishes were kept in refrigerator at 4 • C for 2 h to allow the diffusion of nanoparticles from the corresponding thin coating in the agar and after that cultivated for 24-48 h at 37 • C. The formed sterile zones around the samples were measured in mm.

Cell Growth Inhibition (Most Probable Number of Surviving Cells Test)
The Most Probable Number (MPN) of surviving cells [49] was used to evaluate the inhibition of bacterial growth above the coated samples, placed in Corning ® Costar ® (Northeim, Germany) TC-Treated, 6 well polystyrene plates. The cultivation of bacteria was performed in organic nutrient medium (NB, Conda, Spain) that is most suitable for bacterial growth and multiplication, to reliably prove the antibacterial effect. The concentration of the bacterial cells was 10 5 -10 7 CFU·mL −1 .
OD of bacterial suspension, in absence (control) and in presence of coated sample, was measured every hour (λ = 610 nm, Spekol-11, Carl Zeiss, Jena, Germany). For the MPN test, 10-fold dilutions were made in intervals of 2 h. Surviving cells were determined at different time points as described earlier [50]. All results are average of 3 measurements.

SEM Observation of E. coli on Coated Samples
E. coli on TiO 2 /Ag/Cu coated glass samples were observed after incubation in nutrient medium for 6 h, followed by distilled water rinsing and drying for 12 h under room conditions. Cathode sputtering of gold coating under vacuum in argon atmosphere was made before the SEM observation (Joel JSM 5510, Tokio, Japan).

Results
To evaluate the impact of both Ag and Cu content on the antibacterial activity of the TiO 2 /Ag/Cu composite coatings, five coated samples were selected for this investigation with varied Cu content at almost equal Ag content (Samples 1, 2, and 3) or with varied Ag content at almost equal Cu content (Samples 3, 4, and 5), based on EDX analyses. The exact content of Ti, Ag and Cu, in at.%, estimated from the corresponding detailed Ti2p, Ag3d and Cu2p XPS spectra, are presented in Table 1. Surface characterization of the studied coatings was carried out beforehand to evaluate their antimicrobial activity since the microbial adhesion depends on surface characteristics like surface chemistry and topography, hydrophilic/hydrophobic balance, surface energy, etc. [51][52][53][54][55][56][57].

Elemental Composition (XPS Analysis)
In the survey XPS spectra of all studied TiO 2 /Ag/Cu coated samples, the characteristic peaks of the elements C, O, Ti, Si and Ag were presented; the elements O, Ti, Ag and Cu originating from the cosputtered TiO 2 /Ag/Cu coatings and carbon, originating from the air. The results of the quantitative elemental analysis for Ti, Ag and Cu presenting in the magnetron cosputtered TiO 2 /Ag/Cu thin composite coatings, based on detailed Ti2p, Ag3d and Cu2p XPS spectra, are presented in Table 1.
It is evident that samples 1, 2 and 3 contain almost equal Ag amounts of about 14 at.%, whereas the Cu content varies between 13 at.% and 21 at.%; samples 3, 4, and 5 contain almost equal Cu amounts of 21 at.%-22 at.%, whereas the Ag content varies between 14 at.% and 23 at.%. These five samples were used in the further investigations.

Elements Distribution (EDX)
The elements' distribution is seen in Figure 1, which presents the EDX spectrum (a), an integrated map (b) and the individual elements' distribution (Ti, Ag, and Cu (c, d, e, respectively)) of coated glass Sample 5-TiO 2 /Ag 23.1 at.% /Cu 21.6 at.% . It is evident that all three elements (Ti, Ag and Cu) form chainlike aggregates ( Figure  1c-e), specifically integrated in the TiO2/Ag23.1 at.%/Cu21.6 at.% coating to form dark depression wells, visible in the integrated map, Figure 1b. This picture is characteristic for all studied TiO2/Ag/Cu coatings, including Samples 1-4, and therefore their maps are not presented It is evident that all three elements (Ti, Ag and Cu) form chainlike aggregates (Figure 1c-e), specifically integrated in the TiO 2 /Ag 23.1 at.% /Cu 21.6 at.% coating to form dark depression wells, visible in the integrated map, Figure 1b. This picture is characteristic for all studied TiO 2 /Ag/Cu coatings, including Samples 1-4, and therefore their maps are not presented here.

Surface Morphology (SEM)
The typical surface morphology of TiO 2 /Ag 23.1 at.% /Cu 21.6 at.% coated Sample 5 ( Figure 2b) and that of TiO 2 coated one (Figure 2a), for comparison, is presented in Figure 2. The TiO 2 coated surface is flat (Figure 2a) whereas that of TiO 2 /Ag 23.1 at.% /Cu 21.6 at.% coated Sample 5 (b) characterizes by arrangements of chaotically distributed chainlike aggregates of Ag/Cu particles with different shapes and sizes, most of them submicron, which form dark walls, easily visible in Figure 1b. Such Ag and Cu particles' distribution is characteristic for all TiO 2 /Ag/Cu magnetron cosputtered coatings. Therefore, their pictures are not presented here.

Water Contact Angle, WCA, Surface Energy, E and its Polar, E p and Disperse, E d Parts
As is evident from Table 2, the studied samples demonstrate slight deviations in the WCA (of 63.0-69.1 • ) and the surface energy, E (of 34.96 mN/m-39.21 mN/m), remaining in the region of moderate hydrophilic surfaces. This indicates that the insignificant impact of the surface hydrophilic/hydrophobic balance and surface energy could be expected on the initial microbial adhesion [58]. The diffusion test was employed for quick antibacterial screening of the studied TiO 2 /Ag/Cu magnetron cosputtered coatings. The results are presented in Table 3. Table 3. Antimicrobial activity toward Gram-negative and Gram-positive microbial strains (as a sterile zone, mm; by agar diffusion test) of TiO 2 /Ag/Cu magnetron cosputtered thin coatings depending on the Cu-content at almost equal Ag content (Samples 1, 2, and 3) and depending on Ag content at almost equal Cu content (Samples 3, 4, and 5). The data in Table 3 demonstrate a specific concentration of both the Ag-and Cudependent antimicrobial activity of the studied TiO2/Ag/Cu composite magnetron cosputtered coatings toward different microbial strains. The bactericidal activity of the test microbial species: (E. coli, P. aeruginosa, S. epidermidis and S. holeresius) is different on coatings with the same composition (Table 3, the sterile zones' data in the rows). The sterile zones are different also for the test bacteria on the TiO 2 /Ag/Cu composite coatings containing different amounts of both Ag and Cu (Table 3, the sterile zones' data in the columns). The inhibition zones of the studied bacteria are increasingly enlarged with the increasing Ag content (compare Samples 3, 4 and 5) and less enlarged with the increasing Cu content (compare Samples 1, 2 and 3), which indicates a stronger bactericidal effect of the Ag NPs as compared to that of the Cu NPs. The largest sterile zones appeared for all microbial strains in Sample 5-TiO 2 /Ag 23.1 at.% /Cu 21.6 at.% -the coating with the highest content of both Ag and Cu (Table 3, the last row).

Cell Growth Inhibition
Two microbial strains, Gram-negative, E. coli and Gram-positive, S. epidermidis were used to study the cell growth inhibition on the studied TiO 2 /Ag/Cu cosputtered composition coatings with different Cu content and almost equal Ag content as well as with different Ag content at almost equal Cu content by MPN of the surviving cells. The dynamic of the E. coli growth inhibition is presented in Figure 3.     Table 3. The MPN of the surviving cells and the diffusion assay demonstrate good congruence and increased inhibitory effect against E. coli with the increase in the concentration. Figure 4 represents the inhibitory effect of the same TiO 2 /Ag/Cu composite coatings toward S. epidermidis.     at.% (curve 3); to 18.6 at.% (curve 4) and 23.1 at.% (curve 5) in the studied TiO 2 /Ag/Cu composite coatings at almost equal to Cu NPs content at 22 at.%-23 at.%, increases their activity against S. epidermidis, as indicated by the positions of the curves 3, 4 and 5 in Figure 4. The inhibitory effect toward S. epidermidis cell is most strongly expressed at Sample 5, i.e., at the maximal amounts of both Ag and Cu in the TiO 2 /Ag/Cu coating. The results of the MPN of surviving cells estimation confirm the findings obtained by the quick agar diffusion test (Table 3): higher sensitivity of S. epidermidis than E. coli to Ag NPs and Cu NPs dependent on the Ag and Cu concentration, and antimicrobial activity of the studied TiO 2 /Ag/Cu composite coatings.
Another demonstration of the specific antimicrobial activity of the studied TiO 2 /Ag/Cu composite coatings toward different bacterial species is evident in Figure 5 The significant and specific antimicrobial activity of this coating toward large spectrum Gram-positive and Gram-negative microbial strains is clearly demonstrated by curves 1, 2, 3, 4, and 5. They show that the sharp decrease in the MPN of the surviving cells happens at different time intervals and it is down to different leaving cells number, both depending on the type of the microbial strain. The most sensitive to Ag/Cu as an antimicrobial agent is S. epidermidis, followed by B. cereus, P. putida, P. aeruginosa and E. coli.

E. Coli Cells on Coated Glass Samples
Significant damage to and killing of E. coli, after growth on TiO2/Ag/Cu coated glass samples, were observed by SEM. Typical images are presented in Figure 6. Normally, the E. coli cells look like ovals with size of 0.8-1 μm. They seem to be damaged on the TiO2/Ag/Cu coated samples: deformed, with elongated original shape,  The significant and specific antimicrobial activity of this coating toward large spectrum Gram-positive and Gram-negative microbial strains is clearly demonstrated by curves 1, 2, 3, 4, and 5. They show that the sharp decrease in the MPN of the surviving cells happens at different time intervals and it is down to different leaving cells number, both depending on the type of the microbial strain. The most sensitive to Ag/Cu as an antimicrobial agent is S. epidermidis, followed by B. cereus, P. putida, P. aeruginosa and E. coli.

E. coli Cells on Coated Glass Samples
Significant damage to and killing of E. coli, after growth on TiO 2 /Ag/Cu coated glass samples, were observed by SEM. Typical images are presented in Figure 6. The significant and specific antimicrobial activity of this coating toward large spectrum Gram-positive and Gram-negative microbial strains is clearly demonstrated by curves 1, 2, 3, 4, and 5. They show that the sharp decrease in the MPN of the surviving cells happens at different time intervals and it is down to different leaving cells number, both depending on the type of the microbial strain. The most sensitive to Ag/Cu as an antimicrobial agent is S. epidermidis, followed by B. cereus, P. putida, P. aeruginosa and E. coli.

E. Coli Cells on Coated Glass Samples
Significant damage to and killing of E. coli, after growth on TiO2/Ag/Cu coated glass samples, were observed by SEM. Typical images are presented in Figure 6. Normally, the E. coli cells look like ovals with size of 0.8-1 μm. They seem to be damaged on the TiO2/Ag/Cu coated samples: deformed, with elongated original shape, dehydrated surface and multiple indentations (Figure 6a,b). Large colonies of damaged E. coli cells are also observed in some places (Figure 6b). This could be due to a poured Normally, the E. coli cells look like ovals with size of 0.8-1 µm. They seem to be damaged on the TiO 2 /Ag/Cu coated samples: deformed, with elongated original shape, dehydrated surface and multiple indentations (Figure 6a,b). Large colonies of damaged E. coli cells are also observed in some places (Figure 6b). This could be due to a poured cell content, the last one not seen in these pictures, because it was eluted after the samples had been washed in distilled water. The SEM observation of the E. coli cells on the coated samples demonstrates their contact killing.

Discussion
The surfaces of all TiO 2 /Ag/Cu coated samples (Samples 1-5) are characterized by similar morphology: Ag/Cu chainlike aggregates distributed chaotically in the coating (SEM, Figure 2b), and forming dark wells as visualized by the EDX maps (Figure 1b). All TiO 2 /Ag/Cu coated samples are moderately hydrophilic (WCA of 63.0 • -69.1 • , Table 2). Both the observed surface morphology and the moderate hydrophilicity [58] of the studied coatings suppose good interactions with living cells, i.e., the observed suppression of the test microbial cells growth could not be due to a lack of the conditions for initial attachment (initial adhesion). The suppression of the microbial cells' growth by the studied TiO 2 /Ag/Cu composite coatings and its extent is determined by the chemical composition of the coating, i.e., by the presence of the antimicrobial agent Ag/Cu and its concentration The well expressed cell growth inhibition could be due to cellular interactions with presenting on or/and eluted from the coated surface, Ag and Cu nanoparticles, ions and reactive oxygen species formed as a result of the oxidative processes. It is known that Ag NPs and Ag + , Cu NPs and Cu 2+ are active antibacterial agents, although that the mechanism of their antibacterial action is not fully understood [59]. Our experimental results demonstrate direct contact (SEM observation, Figure 6) and released silver/coper killing (diffusion assay, Table 2; MPN of living cells, Figures 3-5) of the bacterial cells by the studied TiO 2 /Ag/Cu coatings. A similar mode of bactericidal action was proposed earlier for magnetron sputtered TiO 2 coatings, chemically incrusted with Ag NPs [12] and TiO 2 /SiO 2 /Ag magnetron cosputtered coatings [3,17].
There are a number of reports concerning the mechanism of action of Ag and Cu nanoparticles and ions. It was shown that silver ions (Ag + ) interact with disulfide or sulfhydryl groups of enzymes, which leads to the inhibition of metabolic processes [60,61]; silver binds bacterial DNA (deoxyribonucleic acid), inhibiting replication and transcription [62,63] and copper leads to the collapse of some lipopolysaccharide (LPS) patches and alters the permeability and functionality of the outer cell membrane [64]. It was demonstrated that the activity of the NPs is size [27,65] and shape-dependent [66]. The free NPs penetrate inside the bacteria after their attachment to the plasma membrane and interact with sulfur containing proteins and the phosphorus containing DNA, which leads to blockage of the cell division [67]. It is difficult to distinguish the bactericidal activity of the NPs from that of the metallic ions released by the nanoparticles [68]. It was reported that the combined effect of the adhesion and penetration of the NPs and ions is crucial for the biocidal effect, with the plasma membrane being the target of the rapid antibacterial action of the NPs in E. coli [69]. The NPs also interact with the cell envelope of E. coli [70]. Morphological analysis reveals that the hallmarks of cell damage and NP toxicity include an increase in cell surface roughness and a change of the shape and size of the cells [71], such as were observed in our study. The mechanism of the antimicrobial action of Cu NPs and ions is less studied and reports on the mechanism of antimicrobial action of such coexisting with Ag NPs and ions are scarce.
The bactericidal activity of the magnetron cosputtered TiO 2 /Ag/Cu nanocomposite coatings could be due to a damage of the cell membrane and envelope (observed by SEM, Figure 6) usually resulting in a leakage of the cellular content, increased cell surface roughness, changed cell shape and size, and cell death [69][70][71]. The bonding of Ag NPs and Ag + , as well as of Cu NPs and Cu 2+ from the studied coated samples to the microbial cells' proteins, could also contribute, inactivating the electron transport chain and thereby inhibiting the respiration and growth of the cells, as was already reported [24]. The specific antimicrobial action of the studied coatings toward the used test bacterial strains could be explained with the specific features of their cells: rod or coil shape, presence or lack of flagella, cell envelope and membrane, density of the cell membrane, cell wall structure, etc. determining the engulf of the Ag/Cu NPs and ions [72,73].
The in-depth penetration of the mechanism of microbial cells' interactions with Ag NPs and Ag + , Cu NPs and Cu 2+ , as well as with simultaneously presenting Ag NPs/Cu NPs and Ag + /Cu 2+ , needs additional, more in-depth investigations.

Conclusions
This investigation combines Ag and Cu as dopants in RF cosputtered TiO 2 /Ag/Cu functional coatings for medical devices and correlates their characteristics to the antimicrobial activity. The desired thickness, antibacterial activity and its durability can be obtained by the optimization of the operation parameters of the RF cosputtering and an accurately controlled composition of the TiO 2 /Ag/Cu coatings, especially of the Ag and Cu content as dopants. The last one is a convenient tool for the adjustment of the balance between antimicrobial activity and cytotoxicity to the requirements of different medical devices.
The cells' growth inhibition is specific and dependent on the Ag/Cu content for all used test bacterial strains. It is due to direct contact and eluted Ag/Cu-mediated killing, as experimentally demonstrated. The in-depth penetration in the mechanism of the microbial cells interaction with Ag and Cu nanoparticles and ions, as well as with simultaneously presented all of them needs in additional, more in-depth investigations.
It is expected that the studied magnetron cosputtered TiO 2 /Ag/Cu coatings would ensure a broad spectrum of bactericidal activity during the indwelling of the coated medical devices and for at least 12 h after that, with the supposition that the benefits would continue over a longer time.