Innovative PVD Zr–Ti–Ag Thin-Film Metallic Glasses for Dental Applications: Effect of the Silver Content on Antibacterial and Biocompatible Behaviors

Abstract

Zr–Ti based thin-film metallic glasses (TFMGs) constitute promising candidates for biomedical applications. Expected surfaces that are eligible for use inside the body have to gather a high biocompatibility, while preventing bacterial colonization. The objective of this work is to investigate potentialities opened by the innovative ternary Zr–Ti–Ag system. Films were deposited by magnetron cosputtering from independent targets, allowing a wide silver enrichment (10–41 at % Ag). Films were studied regarding their microstructural, mechanical, physicochemical, and biological properties. These properties were discussed in light of the films’ composition. This study established that the structure of the films was strongly influenced by their silver content. The expected amorphous structure was obtained between 20 and 36 at % Ag, for which mechanical properties are high, and compatible with a potential biomedical application. Surface is then very smooth, with presence of dispersed micrometric crystallized islets. Evolution of the surface wettability was found to be primarily governed by the film morphology rather than by its chemical composition. Biologically, a pronounced antibacterial activity against both E. coli and S. gordonii was recorded from 25 at % Ag. Such an antibacterial effect was assigned to silver, as demonstrated by tests carried out with synthetic ions-released solutions. Besides, these films also have evidenced a good cytocompatibility toward gingival fibroblasts. These results highlight the key role of composition-controlled microstructure of Zr–Ti–Ag TFMGs, and demonstrate their potential as antibacterial and biocompatible coatings for dental implant applications.

1. Introduction

Bulk metallic glasses (BMGs) are metallic alloys that solidify into an amorphous structure when cooled rapidly enough to hinder the crystallization process. They have attracted considerable attention due to their unique combination of high strength, large elastic limit, corrosion resistance, and structural homogeneity arising from the absence of long-range order. , However, their processing into large and complex geometries remains difficult because of the massiveness of bulk parts affecting the cooling rates required to avoid crystallization. Moreover, the chemical compositions leading to bulk amorphous structures are complex and multicomponent (4–5 elements in general) to get a high enough glass-forming ability. In contrast, using intrinsically out-of-equilibrium physical vapor deposition (PVD) processes, thin-film metallic glasses (TFMGs) can be deposited owing to their very effective cooling rates (≈10⁹–10¹² K·s^–1). , They indeed strongly limit atomic rearrangement and crystalline nucleation, thereby promoting the formation of fully amorphous structures, even for simple binary systems. Synthesis of metallic glasses by PVD process also enables precise control of composition and thickness, as well as the deposition on complex-shaped substrates.

This compositional flexibility is especially valuable in various fields, especially the biomedical one. The ability to tailor the deposited material’s chemistry indeed allows the incorporation of elements providing specific functionalities, such as enhanced biocompatibility, corrosion resistance, or antibacterial activity. In fact, bacterial infections remain a major challenge in implantology, particularly in dental applications where implants are permanently exposed to a complex oral microbiota. The early formation of bacterial biofilms on implant surfaces can lead to peri-implant infections, compromising osseointegration and long-term clinical success. Preventive strategies aiming to inhibit bacterial adhesion while preserving cytocompatibility are therefore of prime importance. In this context surface properties are critical, as they directly govern protein adsorption, cell adhesion, and bacterial colonization.

Currently, most load-bearing orthopedic and dental implants are manufactured from Ti-6Al-4 V and Co–Cr or Zr–Nb alloys due to their high mechanical strength and long-term stability. , In dental applications, zirconia (ZrO₂) has also gained increasing interest owing to its excellent biocompatibility, corrosion resistance, and favorable soft-tissue response. However, despite their clinical success, conventional implant materials do not inherently prevent bacterial colonization. Consequently, materials that can be used as coatings to tailor the surface properties of existing implants are highly attractive. In this regard, TFMGs have gained increasing attention owing to their compositional flexibility and ability to finely tune surface-related properties, as reported in a recent review.

In the specific context of dental implantology, the transmucosal region at the implant-gingiva interface represents a critical biological junction. This area is continuously exposed to oral microorganisms and constitutes a preferential site for bacterial colonization, potentially initiating peri-implant infections. Simultaneously, stable adhesion and proliferation of gingival fibroblasts are essential to ensure an effective soft-tissue seal. Therefore, thin films designed for this interface must provide a dual role: to confer antibacterial protection on the one hand, while maintaining cytocompatibility with gingival cells on the other hand.

Several Zr- and Ti-based TFMG systems have been extensively studied in recent years as particularly attractive biocompatible coatings for biomedical applications. These systems benefit from the excellent biocompatibility of Zr and Ti, their good corrosion resistance, and mechanical reliability, as demonstrated for Zr–Ti–Si, Zr–Ti–Fe, Zr–Co–Ti, and Ti–Zr–Ta coatings.

The addition of silver is of prime interest for its well-documented broad-spectrum antibacterial activity, already explored for Zr–Cu–Ag system for example. , Incorporating silver into a Zr–Ti matrix could therefore provide a promising strategy to combine the cytocompatibility and corrosion resistance of zirconium- and titanium-based systems, with intrinsic antibacterial functionality given by silver. Despite this potential, the Zr–Ti–Ag ternary system has only recently been explored in TFMG form for antibacterial biomedical devices. −

In this work, we investigate the influence of silver incorporation on Zr–Ti thin films deposited by magnetron sputtering, with a particular emphasis shed on the role of the film’s composition on their structure and biological activity within a dental context. Determining the gingival fibroblast biocompatibility and elucidating the mechanisms behind the antibacterial activity, this study seeks to identify an optimal compositional window that may combine structural stability and multifunctionality for a potential application at the implant-gingiva interface.

2. Experimental Details

2.1. Films Deposition

Zr–Ti–Ag thin films were deposited by magnetron cosputtering of zirconium, titanium, and silver targets in an argon atmosphere. The coatings were grown on (100)-oriented single-crystal silicon substrates with dimensions of 10 × 10 mm². A substrate rotation speed of 15 rpm was maintained throughout the deposition to ensure uniform film composition and thickness, despite the confocal arrangement of the sputtering sources. Prior to deposition, the vacuum chamber was evacuated to a base pressure of approximately 10^–7 mbar using a combination of mechanical and turbomolecular pumps. During sputtering, the argon flow rate was kept at 30 sccm, and working pressure at 0.5 Pa. Zr and Ti were both powered by direct current (DC) power supplies, while Ag was sputtered using a radiofrequency (RF) one. The Zr/Ti atomic ratio was adjusted to approximately 50/50 by tuning the discharge currents applied to the respective DC targets. The silver concentration in the films was controlled by varying the discharge power of the Ag target. All metallic targets had a diameter of 50.8 mm, a thickness of 3 mm, and a chemical purity exceeding 99.9%. The target-to-substrate distance were fixed at 90 mm (for both zirconium and titanium targets) and 120 mm for silver target. The deposition time was adjusted to obtain coatings with a nominal thickness of 1 μm without substrate heating or bias voltage applied during the deposition process.

A series of eight films were deposited by progressively increasing the Ag target power while maintaining an approximately equiatomic Zr/Ti ratio, leading to coatings with increasing Ag contents. The precise composition of each film was determined a posteriori by energy-dispersive X-ray spectroscopy (EDS).

2.2. Structural and Microstructural Characterization

The surface morphology of the films was characterized by using a Zeiss Supra 55 VP field-emission scanning electron microscope (SEM). Some characteristic microstructural defects presenting an ″islet″-like morphology were observed on the surface. Their distribution in terms of both density and size was estimated by image analysis using customized Python scripts (relying on NumPy and tiff file).

The chemical composition of the coatings was determined by Energy-Dispersive X-ray Spectroscopy (EDS) performed at an accelerating voltage of 20 kV. The analyses were carried out using an Oxford Instruments X-Max 80 mm² Silicon Drift Detector equipped with a 60 μm aperture diaphragm and operated with the Aztec software for data acquisition and processing.

To assess the amorphous character of the film, samples were examined by X-ray diffraction (XRD) using a Bruker AXS D8 diffractometer equipped with a Cu Kα radiation source (λ = 0.15406 nm). Measurements were carried out in the Bragg–Brentano θ-2θ configuration.

2.3. Surface Properties

The mechanical behavior of the coatings was characterized by nanoindentation using an Agilent G200 system equipped with a diamond Berkovich indenter. Tests were conducted in continuous stiffness measurement (CSM) mode up to a maximum penetration depth of 500 nm, with a strain rate of 0.05 s^–1, a driving frequency of 45 Hz, and a harmonic oscillation amplitude of 2 nm. Prior to testing, the instrument’s area function for CSM operation was calibrated on a fused silica reference sample (elastic modulus of 72 GPa). For each film composition, at least six indents were performed. The average hardness and Young’s modulus were determined from data corresponding to 9.5–10.5% of the total indentation depth (approximately 47–52 nm).

The wettability of the as-deposited Zr–Ti–Ag coatings was investigated using a Krüss EasyDrop contact angle measuring system. A 2 μL water droplet was carefully placed on the surface with a micropipette. Before measurements, each sample was rinsed with ethanol, dried under a gentle compressed air flow, and left to rest for several minutes to remove any residual moisture. The droplet shape was recorded using a high-resolution Stingray camera, and the contact angle was determined from images taken about 5 s after deposition, ensuring that the droplet had stabilized but not yet evaporated. For every composition, at least 6 measurements were collected under controlled ambient conditions (20.5 °C, 33% relative humidity).

Atomic force microscopy (AFM) characterization was performed using a Bruker Dimension Icon microscope equipped with RTESPA-300 tips operated in tapping mode. All AFM images were acquired with a resolution of 512 × 512 pixels². For each composition, several areas were scanned, and the average surface roughness (Ra) values were extracted from 1 × 1 μm² images to evaluate the surface topology and roughness of the films.

Ionic release measurements were carried out using an inductively coupled plasma mass spectrometer (ICP-MS, Thermo Scientific iCAP TQe). Samples (5 × 5 mm²) deposited on Si substrates were immersed in 1 mL of phosphate-buffered saline (PBS) for 24 h at 37 °C. The experiment was performed in triplicate for each sample. Each solution was acidified with HNO₃ prior to analysis, and five measurements were performed per solution. The concentrations of Zr, Ti, and Ag were quantified. A PBS blank was used as a reference to subtract the background signal.

2.4. Biological Activity

2.4.1. Antibacterial Activity

The antibacterial property of the various films was evaluated using Escherichia coli (ATCC 11303) and Streptococcus gordonii (ATCC 10558) bacteria. E. coli was selected as a standard Gram-negative model bacterium for antibacterial testing due to its widespread use and its involvement in nosocomial infections. S. gordonii, a Gram-positive bacterium, was included to represent bacteria typically found in the oral environment, providing complementary information relevant to potential dental applications.

The bacterial viability was assessed using a colony-forming units per milliliter (CFU/mL) protocol. The tests were conducted under planktonic conditions to simulate the ability of the films to kill bacteria before biofilm formation This approach is particularly relevant for implant applications, especially dental ones, where preventing biofilm formation is essential to avoid infection. Coated samples (5 × 5 mm²) were placed into Eppendorf tubes containing 1 mL of bacterial suspension at an initial density of approximately 10⁶ CFU/mL, prepared in PBS to mimic physiologically relevant conditions. The samples were incubated at 37 °C for 24 h. After incubation, the suspensions were serially diluted and plated on plate count agar (PCA) using the pour plate technique. Following an additional 24-h incubation at 37 °C, colonies were enumerated to quantify the remaining viable bacteria. For all films, experiments were carried out in triplicate to ensure reproducibility. Controls included uncoated silicon substrates and bacterial suspensions without any samples. Statistical differences between samples and the control were assessed using a Student’s t test, after verification of data normality and homogeneity of variances.

Beyond CFU/mL counts, the antibacterial effect of the coatings was also expressed as the antibacterial rate (A.R.), commonly used in the literature, ,− which reflects the percentage reduction in bacterial population compared to reference samples. The A.R. was calculated according to the following eq (eq )­

$$\mathrm{A}.\mathrm{R}.=\frac{\mathrm{Nb}.\mathrm{of\; colonies\; on\; reference}-\mathrm{Nb}.\mathrm{of\; colonies\; on\; sample}}{\mathrm{Nb}.\mathrm{of\; colonies\; on\; reference}}\times 100$$ 1

2.4.2. Identification of Antibacterial Agents

Complementary antibacterial assays were performed using aqueous solutions containing controlled concentrations of Ag, Zr, and Ti to determine the individual contribution to the antibacterial activity of each released metallic ion. Commercial ICP standard solutions (1 g/L) of Ag and Zr prepared in 4 wt % HNO₃, and Ti prepared in ultrapure water, were used as stock solutions. Dilutions were highly diluted in PBS to obtain the targeted elemental concentrations while minimizing the final acid content. For each condition, 1 mL of solution was prepared so as to match the liquid volume used in both antibacterial and ICP-release experiments.

Escherichia coli cultures were prepared by adjusting the optical density of the bacterial suspension to approximately OD ≈ 3.3 measured with a Densimat photometer at 550 nm, followed by a 10-fold dilution. A volume of 10 μL of this bacterial suspension were then added to each 1 mL elemental solution. Samples were incubated for 24 h at 37 °C under static conditions. All experiments were performed in triplicate.

2.4.3. Biocompatibility Evaluation

The biocompatibility of the coatings was assessed using human gingival fibroblasts (Clinisciences 2301-SC) to model cellular responses relevant to dental applications. Samples (5 × 5 mm²) were sterilized in 70% ethanol, rinsed with PBS, and preconditioned in complete fibroblast medium before cell seeding. Cells at passage P2 were seeded at a density of 5000 cells per well in 48-well plates, with glass coverslips used as controls, and cultured for 24 h and 7 days at 37 °C under 5% CO₂. For each composition, two independent samples were analyzed.

At each time point, cells were fixed, permeabilized, and stained to image F-actin (phalloidin) and nuclei (DAPI). Cell morphology and distribution were examined by confocal microscopy, and cell density was quantified from fluorescence images to evaluate proliferation. The cells’ density was quantified from at least nine images’ measurement per sample using confocal microscopy.

3. Results and Discussion

3.1. Characterization of the TFMGs

3.1.1. Composition and Structure

The films were deposited by PVD magnetron sputtering over a wide silver-content range [10–41] at %, while intentionally maintaining a Zr/Ti ratio close to 1 in order to highlight the effect of the silver content on the film properties. Eight films were deposited, and their actual composition was determined by EDS (Table ). As shown in Table , EDS analyses confirmed a gradual increase in the Ag content across the series, with measured concentrations of about 10, 15, 20, 25, 27, 31, 36, and 41 at %. This demonstrated that the progressive increase in the Ag target power effectively led to a controlled compositional gradient. In addition, the Zr/Ti ratio remained close to 1 for all films, as intended.

1. Elemental Composition and Zr/Ti Ratios for the Eight Films, Measured by EDS.

	Zr (at %)	Ti (at %)	Ag (at %)	Zr/Ti
ZrTiAg10	45.7 ± 0.1	43.9 ± 0.1	10.4 ± 0.1	0.995
ZrTiAg15	43.1 ± 0.1	41.3 ± 0.1	15.6 ± 0.1	1.044
ZrTiAg20	40.0 ± 0.2	39.6 ± 0.2	20.4 ± 0.1	1.010
ZrTiAg25	35.8 ± 0.1	39.0 ± 0.1	25.2 ± 0.1	0.918
ZrTiAg27	35.8 ± 0.1	36.4 ± 0.1	27.8 ± 0.1	0.984
ZrTiAg31	34.4 ± 0.1	34.4 ± 0.1	31.2 ± 0.1	1.000
ZrTiAg36	31.9 ± 0.1	32.3 ± 0.1	35.8 ± 0.1	0.988
ZrTiAg41	29.1 ± 0.1	29.5 ± 0.1	41.4 ± 0.1	0.986

To assess the structural state of the deposited films, XRD analyses were performed (Figure ). A clear evolution of the structure is observed with increasing silver content. Films with the lowest silver concentrations (ZrTiAg₁₀ and ZrTiAg₁₅) exhibit sharp diffraction peaks characteristic of crystalline phases, which can be indexed to a hexagonal TiZrAg solid solution, marked by asterisks in Figure . In contrast, samples containing between 20 and 36 at % Ag display only a broad diffuse hump centered between 33° and 43°, indicating the absence of long-range order and the formation of an amorphous structure. The position of the amorphous hump shifts slightly toward higher diffraction angles with increasing Ag content, which can be attributed to a reduction in the average interatomic distance. This trend can be rationalized by considering the metallic radii of the constituent elements, which are 162 pm for Zr, 148 pm for Ti, and 144 pm for Ag. Since Ag has the smallest atomic radius, increasing its concentration is expected to decrease the average interatomic spacing, leading to a more densely packed local structure. At 41 at % Ag, crystalline peaks appears onto the amorphous hump, suggesting crystallization of the film. No diffraction peaks corresponding to crystalline silver are detected, suggesting that Ag atoms may be fully incorporated within the alloy structure. The observed peaks (indicated by asterisks in Figure ) can thus be indexed to a hexagonal TiZrAg solid solution, consistent with the relative metallic radii of Zr, Ti, and Ag and their respective atomic concentrations (29.1, 29.5, and 41.4 at %).

1.

X-ray diffractograms of Zr–Ti–Ag films with Ag contents ranging from 10 to 41 at %. Crystalline films are represented in red and amorphous ones in black. The diffraction peaks corresponding to TiZrAg phases are marked with asterisks (*). Green dashed lines indicate peak positions of fcc phase of silver (pdf file # 04–007–8790).

Considering the intended biomedical and dental applications, the structural stability of the coatings under standard sterilization procedures is a prerequisite. As an example, the ZrTiAg₂₀ film was subjected to both conventional sterilization processes: the γ-irradiation, and the treatment in autoclave. SEM and XRD analyses performed before and after sterilization revealed no detectable modifications in surface morphology or diffraction features, indicating that the film has maintained its structural disordered integrity under both severe conditions (see Supporting Information, Figure SI1).

3.1.2. Surface Morphology of the Films

The surface morphology of the films was investigated by SEM and AFM. The surfaces of each film are shown at two different scales: low magnification in Figure and higher magnification in Figure .

2.

Low-magnification SEM images of the film surfaces. Amorphous samples are outlined in black and crystalline ones are outlined in red (a), AFM topography of the islets for the ZrTiAg₃₆ composition (b), and evolution of islet density and average diameter as a function of silver content (c).

3.

High-magnification SEM images with corresponding AFM insets of the films. Amorphous films are outlined in black, and crystalline ones are outlined in red.

Based on low-magnification surface images, a clear distinction can be made between crystalline and amorphous films, outlined Figure a in red and black, respectively. The crystalline films indeed exhibit a granular surface with asperities, whereas the amorphous films display a dual morphology consisting of a smooth large matrix, in which circular island-like features sometimes emerge.

Such morphological defects have already been mentioned by Zeeman et al. and Borroto et al. in the W–Zr system. − They attributed their formation to growth-induced nucleation within a narrow compositional window, where local stoichiometry and atomic mobility favor partial phase separation. Borroto et al. also demonstrated that the transition from the amorphous to the crystalline state is driven by increasing film thickness, which promotes defect formation in Zr–W and Zr–Mo alloys. −

Characteristic islets developed on the ZrTiAg₃₆ film were characterized by AFM (Figure b). They appeared as 40 nm-high hills emerging from the very flat and smooth surrounding matrix. Besides, their crystalline nature was proved by TEM analyses, which also attested their chemical homogeneity (Figure SI 2). Distribution of islets formed on all amorphous films was quantitatively determined by images analysis (Figure c). It was shown that both the density and dimension of islets strongly varied with the film composition. At 20 at % Ag, islets are highly concentrated, with a density of about 28 000 islets/mm². Their density then sharply dropped, reaching a minimum for the samples containing 25–27 at % Ag, with values between 24 and 4 islets/mm², i.e. approximately 3 orders of magnitude lower than that measured for the 20 at % sample. Above 27 at % Ag, the density increases again, reaching 8145 islets/mm² for the Ag-richest film. Regarding their dimension, the average islet diameter progressively increases with the silver content, from around 1 to 3 μm for 20 and 36 at % Ag, respectively. At 41 at % Ag, the surface morphology appears to result from the coalescence of individual islets where islet boundaries can still be visible, leading to a continuous and roughened structure with visible boundaries between merged regions (Figure a).

Thus, evolution of the islet distribution is linked to the amount of silver in the film. An interesting compositional range around 25–27 at % Ag has been highlighted, for which islets are very scarce and small.

At higher magnification (Figure ), crystalline films, outlined in red, displayed a surface characterized by distinct grains. ZrTiAg₁₀ and ZrTiAg₁₅ revealed a “needle-like” morphology, with elongated grains of approximately 100 nm in length, whereas the Ag-richest film, ZrTiAg₄₁, exhibited a morphology with rounder grains of about 50 nm in diameter. The height of their surface structure morphology was characterized by AFM (insets in Figure ) revealing a relief of about 15 nm in height. In contrast, amorphous films, outlined in black in Figure , exhibited a much smoother matrix without any visible grains, and asperities below 4 nm. All of the films comported small white dots, assigned to nanoscale particles resulting from the deposition step or nanoprecipitates.

3.1.3. Surface Roughness and Wettability

Surface characteristics such as roughness and wettability are key parameters governing the interactions between a material and its biological environment, and are highly sensitive to the material’s microstructure and composition. Indeed, these properties can influence protein adsorption, cell adhesion, and bacterial colonization. ,,,

Figure shows the correlation between the surface roughness (R _a, red curve), the water contact angle values (WCA, black curve) and the silver enrichment in the films. The gray-shaded area corresponds to the amorphous films. The roughness of the different films was measured over 1 × 1 μm² areas, representative of the film matrix and revealed a clear distinction between films containing crystalline islets and fully amorphous ones. Crystalline films exhibited R _a values above 1.9 nm, whereas amorphous films displayed significantly lower roughness (<0.5 nm), consistent with the smoother matrix morphology observed by SEM and AFM in Figure . From a biological standpoint, changes in roughness may modulate protein adsorption and subsequent cell and bacterial adhesion, with rougher surfaces generally promoting higher protein adsorption. Surface roughness, together with surface chemistry, also influences the wettability of the surface. In our case, an inverse correlation between WCA and surface roughness was observed, with lower WCA values for the crystalline films.

4.

Water contact angle (WCA, left axis, black) and surface roughness (R _a, right axis, red). The gray-shaded area highlights the compositional window where the films exhibit an amorphous structure.

The WCA values ranged between 78° and 90°, indicating moderately hydrophilic to near-hydrophobic surface. The WCA first increased from 78° at 10 at % up to 90° at 27 at %, followed by a decrease, down to about 81° at 41 at %. The highest contact angles are measured within the 25–27 at % compositional range, which corresponded to the minimum density of crystalline islets (Figure ). In contrast, compositions exhibiting a higher fraction of crystalline surface features showed lower contact angles. This trend can be explained by the decrease of surface roughness, which modified the effective liquid–solid interfacial interactions and apparent surface energy. ,− It is worth noting that although the absolute variation in WCA (≈10–12°) may be considered as quite moderate, the shift toward higher angles is significant and of prime interest, since it corresponds to the transition from clearly hydrophilic to near-hydrophobic regimes.

The results then showed that the film surface morphology was the main parameter governing their wettability. In this sense, thin-film metallic glasses, almost devoid of nanorough islets (especially in the 25–27 at % Ag range), presented a near-hydrophobic behavior.

3.1.4. Mechanical Properties

The mechanical properties of the Zr–Ti–Ag films were investigated because of their key relevance for biomedical surface applications. In particular, hardness and elastic modulus are important for resistance to wear, deformation, and cracking under physiological stresses, while an appropriate balance between strength and compliance is required to avoid delamination and ensure good adhesion to the substrate. A combination of high hardness with a relatively low elastic modulus is especially desirable, as it may reduce stress shielding by better matching the mechanical response of bone tissue. ,, Zr-based TFMGs are known to combine lower Young’s modulus and a higher hardness than their crystalline counterparts, which makes them mechanically attractive for bone-contacting applications. ,

Nanoindentation results, including Young’s modulus (E), hardness (H), and the derived H/E and H ³/E ² ratios, are presented in Figure . The elastic modulus varied between ∼105 and 118 GPa with Ag content, with values below ∼106 GPa for intermediate Ag compositions (20–31 at % Ag). The hardness showed only weak variations around 4.1 GPa whatever the film composition, confirming published results on Zr-based TFMGs. ,, These combinations of relatively high hardness and moderate modulus are typical of amorphous metallic systems, in which the absence of dislocation-mediated plasticity and the dense atomic packing enhance resistance to indentation. ,,, The distribution of H and E values indicated that the mechanical properties were influenced by both composition and structural state. In particular, the lowest Young’s modulus and highest hardness were measured for films around 25–27 at % Ag, which corresponded to the most homogeneous amorphous structures.

5.

Evolution of the mechanical properties of the Zr–Ti–Ag films as a function of Ag content determined by nanoindentation: (a) Young’s modulus E, (b) hardness H, and (c) H/E (left axis, black) and H³/E² (right axis, red) ratios. Shaded areas and error bars indicate the standard deviation obtained from repeated measurements.

Both H/E and H³/E² ratios, which reflect the resistance to elastic strain and to cracking and plastic deformation, respectively, , exhibit similar trends with increasing Ag content. Both parameters reached their highest values for the 20–36 at % Ag films (about 0.04 and 0.007 GPa respectively), indicating enhanced mechanical performance and improved elastic recovery of the amorphous coatings. For a comparison purpose, conventional biomedical alloys such as 316L stainless steel, Ti-6Al-4 V, and Co–Cr–Mo typically exhibit H/E ratios of ∼0.016, ∼0.04, and ∼0.03, and H³/E² ratios of ∼1.6 × 10^–3, ∼7.5 × 10^–3, and ∼4.5 × 10^–3, respectively. ,, The TFMGs therefore exhibited H/E and H³/E² values in the same order of magnitude as those of Ti-6Al-4 V and Co–Cr–Mo alloys commonly used in biomedical applications, and even higher than that of 316L stainless steel. This suggest an enhanced ability to sustain elastic deformation before yielding together with an improved resistance to contact-induced damage, which are particularly relevant for surface-engineered biomedical materials.

3.2. Biological Characterization

3.2.1. Antibacterial Properties

The antibacterial activity of the films was evaluated using the Gram-negative bacteria E. coli and the Gram-positive bacteria Streptococcus gordonii. E. coli was selected as a widely used model organism in material and biological sciences, while S. gordonii was chosen for its dental origin and its relevance in the context of implant-associated biofilms. The latter is recognized as one of the primary colonizers during oral biofilm formation and belongs to the so-called “orange complex,” which plays an intermediary role between early colonizers such as Streptococcus sanguinis and late, more pathogenic species of the “red complex,” including Porphyromonas gingivalis and Treponema denticola. By forming adhesive bridges within developing biofilms, S. Gordonii facilitates the establishment of these pathogenic bacteria. Therefore, testing this species is crucial, as it provides insight into the potential of the coatings to inhibit the early stages of biofilm development on dental implants. ,

Figure shows the bacterial concentration (in CFU/mL values) after 24 h of immersion (left axis) and the corresponding calculated antibacterial rate (A.R., right axis, blue) for each film relative to the silicon reference (gray bars). Figure a corresponds to E. coli, and Figure b to S. gordonii. For both graphs, amorphous films appear in green bars and crystalline ones appear in red.

6.

Antibacterial activity of the Zr–Ti–Ag films against (a) Escherichia coli and (b) Streptococcus gordonii after 24 h of contact. The bars represent the number of viable bacteria expressed as CFU/mL (left axis), while the blue curves correspond to the antibacterial rate (A.R., right axis). Gray bars indicate the reference (Si substrate), red bars correspond to crystalline films, and green-shaded bars represent amorphous coatings with increasing Ag content from dark to light green. Student’s t test was performed compared to Si reference: *p ≤ 0.05 and **p ≤ 0.01.

Regarding E. coli results, no significant antibacterial effect was detected for the moderately silver-enriched films (10 at % to 20 at %) whatever their structure: ordered or amorphous. However, from 25 at %, a more important and significant decrease in CFU/mL is observed, leading to an A.R. of about 75%. Above 27 at % of Ag, the bacterial concentration drastically dropped by about 3 orders of magnitude compared to silicon, giving an antibacterial rate value higher than 99.8%, quite stable for higher silver contents.

It is worth noting that a similar tendency was also observed for S. gordonii (Figure b). No significant effect on bacteria was indeed recorded below 25 at % Ag, whereas a pronounced decrease was measured from 25 at % Ag, where the A.R. reached 65.5%, followed by a further improvement for higher Ag contents, with A.R. values approaching 93%. Again, the crystalline 41 at % Ag film attested to a slightly degraded antibacterial efficiency, which still had an A.R. higher than 91%.

Overall, antibacterial activity against S. gordonii remained slightly lower than that observed for E. coli.This difference was expected and explained by the structural disparities between Gram-positive and Gram-negative bacteria. As a Gram-positive species, S. gordonii possesses a thick peptidoglycan layer that can limit the diffusion of bactericidal species released from the films, thereby reducing their effectiveness. In contrast, E. coli, a Gram-negative bacterium with a thinner peptidoglycan layer, is more susceptible to the antibacterial action of the films. ,

3.2.2. Ionic Release

To determine whether a correlation did exist between the antibacterial activity of the films and their chemical composition, the concentration of metallic ions released in the medium was measured by inductively coupled plasma (ICP) analysis. Samples were immersed in PBS for 24 h under static conditions and in the absence of bacteria to reproduce the environment used during antibacterial tests. The concentrations of Zr, Ti, and Ag released after 24 h are reported in Figure .

7.

Concentrations of Zr (dark gray), Ti (light gray), and Ag (blue) released after a 24 h immersion of the films in PBS, measured by ICP-MS as a function of silver content in the films, left y-axis. Green curves represent the A.R. considering E. coli (plain line) and S. Gordonii (dotted line), right y-axis.

Globally, both zirconium and titanium were released in higher amounts than those of silver. This trend roughly reflected the elemental proportions of the metals added in the coating. Two regimes can be found for the Ti and Zr ion release below and above 25 at % Ag in the films: below 25 at % concentrations were all close to 10 ppb, and above this critical value of 25 at %, the release of Zr and Ti was more than 1 order of magnitude higher. Concerning silver, the release is more progressive, even if the film containing 27 at % Ag again evidenced an accelerated propensity for the film dissolution.

The general increase in ionic release with the Ag content suggested that higher silver incorporation promoted a more permeable film structure, thereby facilitating ion diffusion into the surrounding medium. This behavior can be rationalized by the formation of a passive surface layer, mainly composed of ZrO₂ and TiO₂. Such oxide layers are indeed known to govern corrosion resistance and therefore control ionic release. Zhang et al. recently reported that moderate Ag additions in Zr–Ti–Ag TFMGs enhanced atomic-scale disorder and promoted dense, noncolumnar microstructures, thereby improving passive-film stability and corrosion resistance. The nearly constant Zr and Ti concentrations measured for moderate Ag contents (10–25 at %) would likely reflect a baseline release controlled by the formation of the ZrO₂/TiO₂-rich passive surface film, through which silver may progressively diffuse toward the liquid environment and oxidize, as reported by Zhang et al. in their study of Zr–Ti–Ag TFMGs. Above 27 at % Ag, the sharp increase in Zr and Ti release would suggest the damage of this protective oxide layer, which would be consistent with chloride-induced breakdown and localized corrosion processes visible for high Ag contents, as already reported in the literature. , Such passive film deterioration would expose the underlying metallic matrix, which would result in an enhanced dissolution of the three constitutive metallic elements.

Overall, the increase in released ions broadly followed the antibacterial trends observed for the films. Indeed, the A.R. values measured for both E. coli and S. gordonii (green curves in Figure ) increased with the ionic release and seemed more closely correlated with the silver release trend, suggesting that Ag⁺ would be a major contributor to the antibacterial character. Silver is indeed known for its antibacterial property, whereas Zr and Ti are usually considered to be biocompatible. ,,, However, contributions from Zr and Ti ions cannot be fully excluded at this stage, and a further series of experiments were performed to definitively identify the respective action of each metallic ion on the antibacterial behavior.

3.2.3. Origin and Mechanism of the Antibacterial Behavior

To identify which released species would be responsible for the antibacterial effect, complementary antibacterial tests were performed using aqueous solutions with controlled elemental concentrations. Synthetic Ag⁺ aqueous solutions were prepared covering a wide range of concentrations. The aim was to identify an intermediate concentration inducing a partial antibacterial activity that could highlight potential synergistic or antagonistic effects with the two other metals. A range of Ag concentrations (1–50 ppb), consistent with the order of magnitude measured in release experiments, was first investigated, revealing a quasi-linear decrease in bacterial concentration from 10⁵ CFU/mL to 0 (Figure SI 3). On the basis of these results, an intermediate concentration of 17 ppb was selected, as it lies between the known 10 and 25 ppb data points, ensuring a partial antibacterial response.

Subsequent tests were performed at this intermediate level using elemental proportions representative of the 27 at % Ag film, previously identified as bactericide (Zr: 50 ppb, Ti: 123 ppb, Ag: 17 ppb), in order to evaluate potential interactions between elements (Figure ). It appeared that bacterial development was not affected when exposed to Zr- and Ti-containing solutions. In contrast, it drastically dropped in contact with all solutions containing silver ions. Therefore, it can be concluded that no antibacterial effect was reported with either Zr or Ti alone. Besides, no synergistic or antagonistic effects were observed in mixed solutions, demonstrating that the antibacterial activity originated solely from silver ions.

8.

E. coli bacterial concentration (in CFU/mL) after exposure to individual elements and their combinations at Zr: 50 ppb, Ti: 123 ppb, Ag: 17 ppb. The reference solution is PBS without any added elements.

These results are consistent with the well-established antibacterial activity of silver ions through multiple simultaneous mechanisms. In fact, they can bind to thiol (-SH) groups in membrane proteins, leading to protein denaturation and disruption of membrane permeability.This interaction compromises the integrity of the bacterial cell envelope, resulting in leakage of cellular contents and loss of vital ion gradients. Moreover, Ag⁺ ions can penetrate inside the cell, where they interact with DNA and ribosomes, inhibiting DNA replication and protein synthesis. They can also promote the generation of reactive oxygen species (ROS), causing oxidative stress and further damage to cellular components, such as lipids and enzymes. Altogether, these effects lead to the inactivation and eventual death of bacterial cells. ,

3.2.4. Biocompatibility Assay

Targeting a use of Zr–Ti–Ag TFMGs inside the human body, the evaluation of their biocompatible character was a prerequisite. Therefore, cytocompatibility of films was assessed using gingival fibroblasts in view of their dental-related applications. Cells were brought into contact with the samples for 24 h (Figure a) and days (Figure b) to evaluate both early stage adhesion and long-term behavior (proliferation) using confocal imaging. For each film’s composition, cell density (cells/mm²) and morphological analysis were used as a quantitative and qualitative descriptors of cell adhesion and proliferation, providing a first indication of cytocompatibility.

9.

Cell density of gingival fibroblasts after 24 h (a) and 7 days (b) of incubation. Gray bars correspond to the glass reference, red ones to crystalline films and blue shades correspond to amorphous films. Confocal fluorescence images of gingival fibroblasts cultured on glass reference and selected ZrTiAg₂₇ films after 24 h (first line) and 7 days (second line) of incubation (c). Cell nuclei are stained in blue and the cytoplasm in orange. Higher-magnification view highlighting filopodial extensions at early adhesion stages (d).

After 24 h of contact (Figure a), cells exhibited good adhesion on most surfaces, with cell densities ranging from approximately 55 to 45 cells/mm², compared with about 65 cells/mm² for the glass reference. A slight decrease in cell number was observed with increasing Ag content up to 25–27 at %, followed by a subsequent increase at higher Ag concentrations. For the film containing 41 at % Ag, the surface was too severely corroded to allow a reliable cell observation. Overall, this evolution appeared to follow the same trend as wettability and surface roughness. Indeed, cell adhesion and proliferation are generally favored on hydrophilic surfaces rather than on hydrophobic ones, such correlations between wettability and fibroblast adhesion have been widely reported in the literature. − This correlation may explain the reduced cell density observed for the 25 at % and 27 at % Ag samples, which exhibited the highest WCA values (around 90°).

For longer times, after 7 days of contact (Figure b), cell proliferation on most samples reached approximately 1700 cells/mm², i.e., about 30 times higher than after 24 h, and was comparable to that observed on the glass reference, indicating that the initial lower cell development at early times was compensated after prolonged incubation. For the film containing 36 at % Ag, two distinct and mutually exclusive behaviors were observed. In some cases, the film remained stable with respect to corrosion during incubation in the cell medium, and a cell density comparable to that measured on the glass control was obtained, indicating good cytocompatibility. In contrast, in other cases, the film underwent severe and widespread corrosion, leading to a fully degraded surface. Under these conditions, no adherent cells were observed, likely due to cell detachment or death induced by severe corrosion, preventing reliable quantification. This all-or-nothing behavior is on the basis of the large variability and increased standard deviation observed for the 36 at % Ag sample in Figure b.

In addition to the quantitative cells’ distribution measurement, their morphology was also considered by confocal microscopy after nuclear and cytoplasmic staining. As an example, Figure c,d presents characteristic images of cells after 1 day (Figure c, first line) and 1 week (Figure c, second line) for the glass reference and the 27 at % Ag-containing film. Images for all other films are provided in the Supporting Information (Figure SI 4). After 24 h of contact, for each composition in the 10–36 at % range, fibroblasts appeared as individual and well-separated cells with clearly visible nuclei (blue) and a well-developed actin stress fibers (orange). Cells displayed an elongated and well-spread morphology with organized actin structures and numerous thin cytoplasmic protrusions extending from the cell body (lateral dimensions within the 50–150 μm range), which reflected well-adherent fibroblasts. This filopodial extension (showed by white arrows in Figure d) also indicated active exploration of the film surface, and robust early stage cell-surface interactions.

After 7 days of incubation, the cells formed a dense and nearly continuous layer covering most of the film surface, compensating for the lower distribution observed after 24 h of contact. At this stage, individual cell boundaries became less distinguishable due to beneficial confluence, reflecting sustained proliferation and surface colonization comparable to that observed on the glass reference.

To conclude, morphological observations are consistent with the quantitative trends deduced from the cells’ density, further supporting the good cytocompatibility of the coatings. Overall, these results demonstrated that the Zr–Ti–Ag coatings evidenced very good cytocompatibility within a compositional window, where silver release remained controlled at a cytocompatible level, providing a pronounced antibacterial effect.

4. Conclusions

This study has demonstrated that silver incorporation into magnetron-sputtered Zr–Ti thin films enabled the control of microstructure and multifunctional properties relevant for biomedical coatings. The Ag content was found to govern the amorphous-to-crystalline transition. Films presented the expected amorphous structure within the [20–36 at % Ag] compositional range. Besides, this disordered structure was resistant to sterilization processes.

Surface properties were found to be primarily governed by the surface microstructure rather than by the surface chemistry. Surface roughness indeed exhibited two distinct regimes, with very smooth surfaces for amorphous films and higher values for crystalline films. The surface wettability was correlated with the surface morphology, particularly with the distribution of micrometric crystalline islet-like growth defects.

A pronounced antibacterial activity was achieved above 25 at % Ag against both Gram bacteria: Escherichia coli and Streptococcus gordonii, the latter being representative of early oral colonizing bacteria. The antibacterial effect was demonstrated to be exclusively associated with the Ag⁺ release into the medium, without any effect of the other metals Zr and Ti. If the thin-film metallic glasses were identified as harmful for bacteria, they nevertheless did not affect the biocompatibility, as they allowed the adhesion and the proliferation of primary gingival fibroblasts, cells maintaining the gum-tooth interface.

Overall, the 25–31 at % Ag-containing TFMGs provided an optimal balance among structural stability, mechanical reliability, and biological performance, defining a new relevant design window for dental coatings. These findings highlighted the potential of Zr–Ti–Ag films as multifunctional antibacterial and biocompatible coatings for dental implants, where limiting early bacterial colonization at the implant-gingival interface is critical to preventing biofilm formation and peri-implant complications.

Glossary

Abbreviations

AFM

atomic force microscopy

A.R.

antibacterial rate

BMG

bulk metallic glass

CFU/mL

colony-forming units per milliliter

CSM

continuous stiffness measurement

DAPI

4′,6-diamidino-2-phenylindole

DC

direct current

EDS

energy-dispersive X-ray spectroscopy

fcc

face-centered cubic

FIB-TEM

focused ion beam transmission electron microscopy

ICP

inductively coupled plasma

ICP-MS

inductively coupled plasma mass spectrometry

OD

optical density

PBS

phosphate-buffered saline

PCA

plate count agar

PVD

physical vapor deposition

Ra

average surface roughness

RF

radiofrequency

ROS

reactive oxygen species

SEM

scanning electron microscope

TEM

transmission electron microscopy

TFMG

thin-film metallic glass

WCA

water contact angle

XRD

X-ray diffraction

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsomega.6c05810.

• Sterilization investigations (figure and description); FIB-TEM cross-section of ZrTiAg₂₀ (figure and description); bacterial concentration as a function of Ag⁺ (figure); and cell confocal fluorescence images at 24 h and 7 days of incubation for all films (figure) (PDF)

N.L.: Investigation, data analysis, data curation, visualization, writing of the original draft, reviewing, and editing. M.R. and C.D.L.: Supervision of the biocompatible and bacterial parts. P.S.: Conceptualization, funding acquisition, reviewing, and editing. H.B.: Investigation (AFM experiments), data curation, and reviewing. D.P. and L.B.: Experiments (films deposition and XRD, biocompatibility). V.D.F., A.G., N.C., A.L.C.L., L.P., F.B., J.P.C., F.G., J.F.P.: Reviewing.

The authors declare no competing financial interest.